Skip to main content

Full text of "R and D Productivity: New Challenges for the US Space Program"

M ■ 

-n '*v ' \'«^ *.-^» - 4 

I^/f/9 /jd/^AJ^aJ 







for the U.S. Space Program 


of Houslon 

Clear L^ahe 

A, C9>i!« spons o red by the Lyndon B. Johnson 
3 —m jnd the Ulnn«.sJty of Houston-dear 
^ cok^C^iation wnth the American Institute of 
AarooKn^s and Astrorsutics and the American 
ft<JCii»c>«-:ty CenKr. 

Gi/f 1 15635 




W«»A.W-.«--l. - ^K 





Edited by 

Otis W. Baskin 

Center for Advanced Management Programs 

University of Houston-Clear Lake 


Leslie J. Sullivan 
NASA/Johnson Space Center 

A Conference Sponsored by: NASA/Johnson Space Center 

and the University of Houston-Clear Lake in cooperation 

with AIAA and the American Productivity Center 

Houston. Texas 

September 10, 11, 1985 

*-.'jr5*<V' - '' 


u>'^:'^%^^\.;'- ■-■■"" ■ ■ ■ (t> 

Copyright © 1^85 by 

University of Houston-Clear Lake 

This work relates to NASA Contract No. 
JSC 3-85-8273. The U. S. Government has 
a royalty-free license to exercise all rights 
under the copyright claimed herein for Government 
purposes. All other rights are reserved. However, 
copyright is not claimed for any portion of this 
book written by a United States Government 
employee as part of his or her official duties. 

■ -J 

-^•T^^^t-.. -' 


JlLHIi'iW^-"** •■ I I 




Central to this nation's lead«.^a>l^T^ i^ 
technologies is our long-term co^Umen? In ^"""'^T ^""^ °'^" advanced 
The historic accompllsh^entrof ?Se Innli p'"^""^ ^""^ development. 
attention of the world but also depone traLd^r""' ^^^ °^'^ "^^""^ '^^ 
great technological breakthrouehrrhrn k ^ "'4^"=^ ""^ achieving 
combined ventures. Now with Sn^^T^^ 8°V"™«nt /industry /university 
future of Space Station close at Lna'"'r"'""" " '^^^^'>' ^^^ ^he ^ 
productivity of the natloVa^R^D o ganlL io^ns a" '' '°'"f^' ^"^" ''^ 
round of major achievements ^n space as w^? ^^^^"^^i^l for the next 
disciplines. ^ " ^^ *'^^^ ^s other industries and 

Of HoSLniat"/!..':'"::: c's::it5:d"L\h"? ^" ?:^«^^°^> ^^« ^-^--^^^ 

their respective role as res^rch and Sv ? '""'"'^ ""' productivity In 
Certainly, the flow of new Ideas nrodn^^I^^r"' "'^S^^l^^tlons. 
to secure a bright future for ou^ wor^d H ' "f ^^^hnologles necessary 
strategies and technique: of managing r,d'"?J/°\"^"' ^"^""- ^° 'he 
conference, we hope to build upon the saUA f ^"^^ sponsorship of this 
experience from the American space nro^°.' °^ ^^'^ """^ P'^^^^"' 
accomplishments In sciencranrtLhnolS^ """"'' '"""'^ 

confer:nc:'j":ir:iu':.::r:o"ei positr'--^ r ^"^--"- at the 

organizations represented W^.^ P°fitive catalyst for all 
cooperating In thirM^t "venture ' '"'' °"^ organizations are 

aid D. Grif 
NASA Lyndon B 
Space Center 



Thomas M, Stauffer 

University of Houston- 
Clear Lake 






This volume conta''.n8 the manuscripts of papers presented as part of 
"R&D Productivity: N^iw Challenges for the U.S. Space Program," 
September 10-11, 1983. This conference was planned and operated as a 
joint project betwten t. e University of Houston-Clear Lake and the 
Lyndon B. Johnson Space Center in cooperation with the American 
Institute for Aeronautics and Astronautics and the American Productivity 

The 51 papers published in this volume were selected for 
presentation through a rigorous review process from 112 papers submitted 
for consideration. However, all those who submitted papers made a 
distinct contribution to the success of this conference whether or not 
Lheir work is contained in this publication. The response to the call 

for papers was overwhelming and many excellent contributions had to be : 

turned down because of the limitations of time and space. Therefore, we 
express our thanks to all those who submitted abstracts because their 
ideas, work and creativity are the foundation of our efforts. 

A conference of this size and scope requires the support of many 
individuals. Thanks are due the 22 reviewers from AIAA, JSC, and UH-CL 
whose tireless reading, rereading, evaluating and organizing produced a 
program high in both quality and interest. Special thanks to Bob Lewis 
of AIAA for his ability to enlist the help of so many colleagues in this 

Special mention also must be made of the vast and various i 

contributions of Alma Martin from JSC. Her ability to facilitate, \ 

coordinate and otherwise cut through red tape made our efforts more 
productive. In addition, the work of Betty McSheehy, Jeane Conway- James 
and Steve Dubuc from the Center for Advanced Management Programs was 
tireless and professionally performed. Their efforts in organizing and 
executing the details of this conference made us efficient and effective. 

Our appreciation must also be expressed to Harry M. Porter and his •, ^ 

staff at the JSC Printing Management Branch. Their ability to meet "• 

tight schedules with quality work has made this book a reality. ' 

Finally, the contributions of Les Sullivan to the concept and \ 

content of this program and the guidance of Wayne Young in the process \ 

are gratefully acknowledged. i 

Otis W. Baskln 




-KJ, rl.ik. .-1. 




R. Wayne Young, Chairman 
Johnson Space Center 

Otis W. Baskln 

University of Houston-Clear Lake 

Robert E. Lewis, President, Houston Chapter AIAA 
and Johnson Space Center 

Alma S. Martin 

Johnson Space Center 

Leslie J. Sullivan 

Johnson Space Center 

Carl Thor 

American Productivity Center 


Betty McSheehy - University of Houston-Clear Lake 
Jeane Conway-James - University of Houston-Clear Lake 
Steve Dubuc - University of Houston-Clear Lake 


A- ; 


Forward ill 

Preface Iv 

Acknowledgements v 


Implementing Quality/Productivity Improvement Initiatives In an 

Engineering Services Environment, Roger R. Ruda, 1 

The USAF Systems Command and R&D Productivity, 

Vince Luchsinger, 10 

Self Renewal: A Strategy For Quality and Productivity 

Improvement, Don Hutchinson, 18 


*' The Key to Successful Management of STS Operations: An 

^ • Integrated Production Planning System, William A. Johnson 

/ ' and Christian T. Thomasen, 33 

The Star System: A Production Engineering Approach to STS, 

Robert C. Angler and Joe H. Wilson, 41 


Management Behavior, Group Climate and Performance Appraisal f 

r at NASA, George Manderllnk, Lawrence Clark, William Bernstein, \ 

and W. Warner Burke, 52 

Mentoring as a Communication Channel: Implications for 
Innovation and Productivity, Robert W. Boozer and Lee 
Avant, 63 



'■' ^ 

Managing Cooperative Research and Development Ventures, '^ 

William J. Murphy, 76 ' 
Government to Government Cooperation in Space Station 

Development, Sam H. Nassiff, 88 ' 

Productivity Issues at Organizational Interfaces, j 
A. W. Holland, 109 


Efficiency and Innovation: Steps Toward Collaborative 
. Interactions, Cynthia A. Lengnick-Hall and Dottsld King, 122 

The Gamma Ray Observatory Productivity Showcase, Richard 
r' L. Davis and D. A. Molgaard, 132 

A Case Study in R&D Productivity: Helping the Program 
-I Manager Cope With Job Stress and Improve Communication 

I Effectiveness, Wayne D. Bodenstelner and Edwin A. Gerloff , 142 


' vl 

W W JtbJlX^il^. -A. .^ 


Technical and Management Information System; The Tool 
I . For Professional Productivity on the Space Station 

Program Office, G. Montoya and P. Boldon, 150 
New Technology Implementation: The Inter-Actlon of 
Technical, Economic and Political Factors, James W. Dean, Jr., 
Gerald I. Susman and Ptunela S. Porter, 165 


Computer Costs: A Method to Compare System Costs Versus 

Costs Incurred by Users Due to Inadequate Computer Resources, 
; E. T. Crucian, 177 

Office Automation: The Administrative Window Into the 
^i Integrated DBMS, Georgia H. Brock, 192 

Improving Decision Processes of Management Through Centralized 
"i CoEununlcatlon Linkages, Don F. Slmanton and John R. Garman, 202 



l\ A Performance Measurement System For Engineering Services, 

'] Richard L. West, 214 


i Some Key Considerations In Evolving a Computer Systems and 

Software Engineering Support Environment For the Space Station, 

Rodney L. Bown and Charles W. McKay, 222 

A ProducLivity Improvement Project In a Design Technology 

Group, R. M. Nordlund, S. T. Vogt, A. K. Woo, 231 

Improving Engineering Effectiveness, Janet D. Flero, 239 

r Software Productivity Improvement Through Software 

Engineering Technology, Frank E. McGarry, 249 


Improved Productivity Through Interactive Communication, 

Phillip P. Marino, 266 

Soclo-Technlcal Integration of the Work Place, 

George L. Carter, 275 

Information Technology, Rolf T. Wlgand, 289 

Training Managers For High Productivity, Robert M. Ranftl, 310 


Increasing Productivity in Flight With Voice Commanding, 
J William T. Jordan, 326 

Group Structure and Group Process for Effective Space 
Station Astronaut Teams, John Nicholas, 330 
Space Crew Productivity: The Driving Factor In Space 
Station Design, Harry L. Wolbers, 340 


- «--l()!»«i<'(V' -' 


^, At-l'>»,V',x 



The Space Station and Human Productivity; An Agenda For 

Research, Claudia Bird Schoonhoven, 352 

Post-IOC Space Station: Models of Operation and Their 

Implications for Organizational Behavior, Performance 

and Eff.ectlveness, Scott Danford, James Meindl and Raymond 

Hunt, 368 


The Role of Artifical Intelligence and Expert Systems in 
Increasing STS Operations Productivity, Chris Culbert, 378 
Application of Modern Tools and Techniques to Maximize 
Engineering Productivity in the Development of Orbital 
Operations Plans, J. S. Manford and G. R. Bennett, 383 
Onorbit Mission Planning Using the Shuttle Trajectory 
and Launch Window Expert System, Peter R. Ahlf , 394 
Automated Crew Procedures Maintenance, 
P. C. Holllngshead, 404 


Streamlining: Reducing Costs and Increasing STS Operations 

Effectiveness, R. K. Petersburg, 416 

Actions For Productivity Improvement in Crew Training, 

Gerald E, Miller, 425 

Ground Processing of the McDonnell-Douglas Payload 

Assist Module (PAM), C. E. Bryan and D. A. Maclean, 437 

Space Shuttle Descent Design - From Development to 

Operations, R. E. Kite, III and T. J. Crull, 450 


Productivity Increase Through Implementation of CAD/CAE 

Workstation, Linda K. Bromley, 462 

R&D Productivity Improvement at Hone^'well: A Case Study, 

William E. Lyons, 474 

Increasing Productivity of the MCAUTO CAD/CAE System By 

User-Specific Applications Programming, Susan M. Piotrowski 

and Tom H. Vu, 481 


Quality Circles - Organizational Adaptations, Improvements 
and Results, Ralph Tortorlch, 490 

Productivity Enhancement Planning Using Participative 
Management Concepts, M. E. White and J. C. Kukla, 495 
Results of Innovative Communication Processes on 
Productivity Gains in a High Technology Environment, 
Brenda J. Kelly, 506 



f ( ..Vil..'*-.Ui,.\' 



White Collar Productivity Improvement: A Success Story, 

Don Hutchinson and E. L. Franren, 518 

White Collar Productivity Improvement in a Government 

Research and Development Administrative Support Organitatlon, 

Bradley J. Baker, 529 

Quality Circles in R&D Organizations: Some Lessons Learned at 

NASA's Kennedy Space Center, Edward J. Heclcer, 539 


Very Large-Scale Integrated Circuits For Switching Data 
Channels, Robert Henrich, 548 

Modular System Design For Space Station Data Handling 
Requirements, M. T. Stowe 560 


Executive Summary and Findings, Steven A. Leth 













Roger R. Ruda, McDonnell Douglas-Houston 


Quality/Productivity Improvement (QPI) initiatives in the 
engineering environment at McDonnell Douglas-Houston include several 
different, distinct activities, each having its own application, yet 
I all targeted toward one common goal - making continuous improvement 

j a way of life. The chief executive and the next .;wo levels of manage- 

•j ment demonstrate their commitment to QPI with hands-on involvement in 

:| several activities. Each is a member of a QPI Council which consists 

I of six panels - Participative Management, Communica ns. Training, 

Performance/Productivity, Human Resources Management and Strategic 
Management. In addition, each manager conducts Workplace Visits and 
"Bosstalks", to enhance communications with employees and to provide 
a forum for the identification of problems - both real and perceived. 

Quality Circles and "Project Teams" are well established within 
McDonnell Douglas as useful and desirable employee involvement teams. 
The continued growth of voluntary membership in the circles program is 
strong evidence of the employee interest and management support that 
have developed within the organization. 

Every member of upper management and over one-third of the 
remainder of the workforce have been trained and are actively involved 
in some activity to enhance McDonnell Douglas' continuous improvement 






In d McDonnpll Douglas Corporate Policy Statement, 
S. N. McDonnell, Chairman and Chief Executive Officer, gave the follow- 
ing direction, "The responsibility for quality and productivity im- 
provement rests with every MDC employee. Corporate, component and 
subsidiary chief executives have primary responsibility for establish- 
ing and maintaining an effective quality/;. roductivity improvement 
process that involves all employees." 

McDonnell Douglas-Houston's Quality/Productivity Improvement 
(QPI) process emphasizes six specific initiatives - Participative 
Management, Communications, Training, Performance/Productivity, Human 
Resources Management and Strategic Management. Continuous improvement 
in each of these areas is the common goal of the activities that have 
been implemented to date. There is still much to be done, but with 
the follo\ving mechanisms and activities in place, and with the total 
commitment of top management to keep them in place, it will get done. 




In January, 1984, the Director and Chief Executive of McDonn'^11 
Douglas' Houston Operations established a Quality/Productivity Improve- 
ment Council (QPIC) chartered to identify and implement activities to 
further the Corporate QPI initiatives. Membership of the council is 
composed of the Director and the next two levels of management. Mem- 
bership is mandatory and presently consists of 32 managers. Each 
manager is also assigned to one of the Council's six panels (Figure 1). 


Figure 1 
Quality/Productivity Improvement Council 














•"^ . '^SSS ^ '":'^' 



Each panel meets weekly and reports progress and recomnendations twice 
monthly at the QPIC meetings. The QPIC is chaired by the Chief Execu- 
tive to emphasize top management commitment and involvement and a new 
management position was created to assist him. The new manage- of 
Quality/Productivity Improvement is assigned full-time responsibility 
for coordinating all QPI activities within the Division, other compon- 
ents and Corporate, and for identifying methods and techniques to 
train and motivate every employee to become Involved in the QPI 

Initial training in quality improvement techniques was pro- 
vided QPIC members utilizing the "Juran on Quality Improvement" video- 
tape series [ 1 ) and the Juran "Project Team" approach was applied by 
each panel. Follow-up training has included "Action Plans for Imple- 
menting Quality and Productivity" [2] and "Toward Excellence" (3 1. 
Agenda items at each regularly scheduled panel and council meeting 
emphasize the commitment to continuous improvement and provide addi- 
tional opportunities for discussing past, present and future activi- 
i ties. 

f This QPI approach has resulted in several improvements in 

1 each initiative area as illustrated by, the following examples. 

Participative Management 

The Participative Management (PM) panel began its activities 
identifying what type of PM activities could be applied in the engi- 
neering organization and defining available training to help under- 
stand and implement PM techniques. Several consultants and much 
library research led the panel to conclude that PM was no single tech- 
nique that could be, or was, packaged, available and applicable in the 
aerospace engineering support services envirp.iment. The panel then 
developed a workshop, "Introduction to Participative Management", and 
conducted it for all employees of the division. The workshop included 
examples of participative management styles, the organizational and 
environmental barriers to using PM techniques, and the advantages and 
effectiveness of employee involvement in management decisions. Pro- 
jects the PM panel is working on now include defining PM training 
requirements for supervisors and defining techniques for evaluating 
mariat^'^rs and supervisors effective use of PM techniques in their work 


There have been noticeable improvements in communications with- 
in McDonnell Douglas as a result of the initiatives described. Shortly 
after Its formation, the Communications panel recommended publishing a 
local newsletter to increase the workforce's knowledge of ongoing Divi- 
sional activities. "The Houstonian", a high quality, eight to twelve 
'- page newsletter is now published monthly. It has greatly Improved top- 

•; to-bottom, and, with employee inputs and articles, bottom- to-top com- 

munications as well . 


"Workplace Visits", or Management by Wandering Around (MBWA), 
is utilized extensively by all QPIC members. Visits are unscheduled, 
informal, and provide direct bottom-to-top communications and problem 

"Bosstalks" are held regularly, usually weekly, by the Divi- 
sion Director. Thirty to forty personnel from all organizational 
levels are invited to attend these meetings. The meetings have no 
fixed format, the Director solicits topics and problems for discussion 
from attendees. Problems identified at these meetings are later 
assigned to managers for resolution. Some typical problems and their 
resolutions are shown in Figure 2. Each Department Manager is en- 
couraged to conduct similar meetings with his personnel. Lateral 
communications is greatly ennanced by having the top three organiza- 
tional management levels in attendance at the regularly scheduled 
QPIC and panel meetings. 

Figure 2 



(<v200 PER MONTH) 
















Training, in all areas - management, administrative and tech- 
nical is needed to ensure a top quality, productive workforce and 
sustain a continuous improvement process. 

Prior to the formulation of the Training panel, various types 
of training were arranged for and scheduled by individual -nanagers. 
The Training panel's initial task was to determine what skills were 
being taught to whom, what other drills were needed and what would be 
needed in the future. 

' ■"^r-ssp---'.' 


a««Ji««i. ^-i •• 


\ A consultant firm was used to help define what the management 

t training needs were for supervisors, mid-level and executive level 

managers. Surveys of project and department managers identified 
technical requirements. After the needs were established, the search 
for effective training programs was begun. Some are being developed 

I ! 

\ An evening study program existed to provide employees an 

J opportunity to develop skills that would contribute to their personal 

growth. The program was greatly expanded and a full time tratnfng 
I coordinator was hired to enhance the programs quality and content. 

Training in statistics and quality control is presently being 
provided at all levels to aid personnel in recognizing what can and 
should be measured or monitored to improve processes, and to detect 
and resolve errors. This training is being conducted as part of a 
"Continuous Improvement" workshop, where the need for statistics is 
examined and examples of their use in monitoring processes are pre- 


^ The major objective of the Performance/Productivity panel has 

"f been to define a Performance Measurement System (PMS) applicable to 

•k an engineering organization where performance is considered to consist 

i of both quality and productivity factors. Their initial efforts were 

f to analyze non-productive engineering time so that measurements of 

* this factor could be an indicator of improvements. The collection 

rf and tracking of information proved to be too unwieldly however, and 

4 this approach was dropped. The PMS presently in use is much more 

f flexible and applicable to the whole organizational environment 14 1. 

t It is based on the American Productivity Center's "family of measures" 

■-I system that was developed by the panel after attending APC's "Know- 
ledge Worker Measurement" workshop. 


H uman Resources Management 

The Human Resources Management panel's initial projects were 
to analyze and improve the employee career counselling procedures, to 
devise a means for making employees more aware of job opportunities 
within the company and to increase the effectiveness and quality of 
the Recognition and Awards program. The new Job Posting System is in 
place, allowing employees to be aware of job openings and apply, con- 
fidentially, for a job that may be more in line with their career goals, 


The present task being worked by this panel is to assess the ''-■'■ 4 

status and results of the Continuous Improvement process at McDonnell "•' 

Douglas-Houston, and to identify areas where further improvements \ 
should be emphasized. 



ii ^s^^^^.:--^ 



The Career Counselling Supervisor's Handbook was developed 
and is in use by each supervisor in the division. Existing Certifi- 
cates of Achievement, New Technology Awards, and the Employee Sugges- 
tion Program were revised and revitalized. The "Directors Award" was 
created to reward the one most outstanding achievement each month. 
The project presently being worked by this panel is to ferret out 
restrictive company policies and procedures and revise them to be 
more in line with the new organizational culture. 

Strategic Management 

The Strategic Management panel was just recently added to the 
QPIC as an individual improvement initiative. The panel's charter is 
to establish strategic management as a "Way of Life", where the 
approach to business is the identification of a) the changes the 
company faces in the future and b) the alternatives that can be de- 
veloped for dealing with those changes. The initial project then, 
for this panel, is to define the long range business plans and the 
current business strategies for implementing those plans. 

Quality Circles 

Quality Circles at McDonnell Douglas-Houston were a forerunner 
of the formal QPI organization. They were established in 1983 with 
the formation of three pilot engineering circles. There was some 
doubt initially as to the applicability of traditional quality circle 
techniques within a purely engineering ?nvironment. At present, there 
are twenty-five circles (23 in engineering) involving over 260 em- 
ployees, and the original three circles are still operating effec- 
tively. Six circles are presently self-facilitating, i.e., they 
totally lead and manage their own activities. 

The circles program at McDonnell Douglas is named SPACE - 
Solving Problems Among Creative Employees. Each circle leader receives 
28 hours of training which includes traditional quality circle problem- 
solving techniques, as well as group leadership techniques. Circles 
are formed within existing work groups, each usually having six to 
fourteen members. Each circle, with the help of a facilitator, pro- 
ceeds with p'-oblem-solving techniques training and immediately begins 
workin ' on a work-related problem of their choice. Circles have vol- 
untary !iiemL3rship and meet for one hour per week on company time. 
Whe, a recommendation for a solution to a problem has been formulated, 
i: presentation is made to management. Management has responsibility 
for making the decisions to implement or not implement suggestions. 
Almost every proposed solution has been implemented to date. Typical 
problems chosen include: improving the contract documentation review 
cycle, improving intragroup communications, standardizing crew train- 
ing scrin';i, reducing memo routing-time and improving Accent quality 
assurar.ce procedures. 



_,^^t'--Vfm.Vtls'- ^C" 

^ ^-^J^^^^v 


A key ingredient to a successful quality circle program is 
managetnent commi tment , and as that has been provided in support of 
QPI activities, the circles have done well in this engineering en- 

Project Teams 

Project Teams are created at McDonnell Douglas whenever a 
specific management-related problem needs to be solved. Membership 
of each team is comprised of management and non-management personnel, 
at any level, across the organization, who have the knowledge neces- 
sary to solve the problem. Each team leader receives group problem- 
solving training, either through the Quality Circles program or 
"Juran on Quality Improvement". An individual team exists only as 
long as it takes to define, analyze and solve that particular problem. 
All QPIC members and eighty-five middle managers in Houston have taken 
the Juran series, and over 260 supervisors and engineers have received 
Circles training. More are being trained continuously to meet our 
goal of 100% employee involvenient in QPI initiatives. 


The Quality/Productivity Improvement Council is the heart, and 
the brains, of the McDonnell Douglas-Houston QPI initiatives. The six 
QPI Council panels - Participative Management, Communications, Train- 
ing, Performance/Productivity, Human Resources Management and Strate- 
gic Management are continually id'-'^tifying, developing and implement- 
ing plans and activities to capit : :2e on every improvement opportunity. 
This top management team, led by Bnl Hayes, Dii*ector and Chief Execu- 
tive of Houston Operations, is committed to continuous improvement. 

The remainder of the workforce is also becoming very involved 
in the continuous improvement process. There are twenty-five Quality 
Circles and the number is continuing to grow. Employee involvement 
and training in QPI techniques presently includes over one-third of 
the workforce, one hundred percent is our goal. Employee recognition 
for achievements provides additional motivation for everyone to get 
involved. Implementation of continuous quality/productivity improve- 
ment activities is well underway at McDonnell Douglas-Houston. 

I ._ i) 





111 Juran, J. M., Juran on Quality Improvement , Juran Enterprises, 
Inc., Second Edition, 1982. 

(21 Tribus, Myron, Action Plans for Implementing Quality and 
Productivity , MIT, 1983. 

131 Peters, Thomas J. end Zenger-Miller, Inc., Toward Excellence , 

141 West, R. L., A Pe rformance Measurement System for Engineering 
Services , Paper presented at NASA-Lyndon B. Johnson Space Center 
conference, "R&D Productivity: New Challenges for the U.S. 
Space Program", 1985. 

/* ..i;,.i5\j^: t?k*v'- <>~'''K.:"''~^- ' ^jy 


Roger R. Ruda, Manager of Quality/Productivity Improvement 
at McDonnell Douglas Technical Services Company-Houston, Texas, re- 
ceived a Bachelor's degree in Electrical Engineering from the 
University of Wisconsin in 1965. Since that time, Mr. Ruda has 
been employed with McDonnell Douglas, where he has held various 
engineering and management positions. He was appointed to his pre- 
sent position in September 1983. 



^^ '' N8 6 -15 159 



Vlnce Luchslnger, Maj.Gen. USAFR 

Mob. Asst. to Commander, USAF Systems Command 

& University of Baltimore 


i The United States Air Force Systems Command (AFSC) is charged 

- ^ with the development and acquisition of aerospace technology systems. 

f Much of that activity is concerned with space systems development, 

> acquisition, and operations. Heavy emphasis Is being placed on produc- 

tivity in organizational and process functions which will keep aerospace 
systems on the leading edge of technology, with plans extending capa- 
bility into the future. The productivity emphasis ranges from people- 
oriented activities to resource and technological functions which support 
national aerospace objectives. The AFSC space-related missions will be 
discussed as a special area of productivity efforts. 


The primary mission of the Air Force Systems Command (AFSC) is 
to advance aerospace technology, apply it to operational aerospace 
systems development and improvement, and acquire qualitatively superior, 
cost-effective, and logistlcally supportable aerospace systems. The 
Command has been Involved in space missions from the earliest United 
States ventures, going back to the early 1950's in exploratory work by 
General Bernard Schriever a past Systems Command coraiiander . 

The Air Force Systems Command also supports the major space 
responsibilities of the Department of Defense, including basic and 
applied research, development, test and engineering of satellites, space 
launches and missions, boosters, space probes, and related space systems. 
The Command is presently heavily involved in the research activities of 
the Space Defense Initiatives program, under the command of a space 
vetern of AFSC experience. General Abrahamson. In addition, AFSC 
supports many NASA programs and projects that operate under agreements 
between the Department of Defense and NASA. 




i^.di-i%' c Jk^riv" 



AFSC Resources 

The Conunand Is one of the larger Air Force Comnands. In scope 
of financial activity, it would rank sixth among the Fortune 500 of 
United States corporations. The personnel ranks of AFSC list 27,500 
military and 29,500 civilian personnel. The nature of its research, 
development, test, and acquisition means that AFSC is the primary 
Air Force employei of scientists, engineers, and technically oiiented 

In the current 1985 fiscal year. Systems Command is managing 
approximately $38 billion. Within that amount, $23.1 billion goes for 
procurement of aircraft ($16 billion), missiles ($5.3 billion), and other 
equipment ($1.8 billion). Beyond that, $10 billion is applied to 
research, development, test, and evaluation (RDT&E), $1.5 billion for 
operation and maintenance, and $540 million for military construction. 
The remaining $3 billion comprises foreign military sales ($1 billion), 
relmbursables ($1.2 billion), and military pay ($900 million). 

The magnitude of the resources entrusted to AFSC dictates a heavy 
responsibility for productivity in discharging its missions to ensure 
the best output for the resources involved. As a vital segment of the 
Air Force structure, AFSC administers thirty-eight percent of the total 
Air Force budget, while utilizing only 6.5 percent of the personnel of 
the Air Force. As a final indicator of the key role of the Systems 
Command, AFSC currently administers 29,000 active contracts valued at 
approximately $108 billion. 


The heart of AFSC contains four -oduct divisions: 

1. The Space Division (Los Angeles AFS) which develops, 
tests, and supervises launch and operation of space craft. 
Facilities include V; ndenburg AFB (which will provide for 
future shuttle launches and recoveries) , the eastern space 
facilities at Patrick AFB, and the Space Technology Center 
at Kirtland AFB. 

2. The Electronics Systems Division, which develops, tests, 
and procures electronic systems for aerospace and ground 

3. The Armament Division which develops and tests con- 
ventional weapons. 

Ac The Aeronautical Systems Division which develops and 
tests aircraft systems. 

The Ballstic Missile Office at Norton AFB is involved in the 
development and testing of missiles and their component systems. The 
Foreign Technology Division monitors a wide variety of foreign 


^ .^.M/'-^MLvt-^ ■^^,: ^<. ■"' " ' ' ■ (^ 


technological activity, with special emphasis on space. The Contract 
Management Division oversees the contracts for acquisition of aerospace 
systems, including space systems. The Aerospace Medical Division 
studies the effects of manned flight, with space flight cons T-itu ting 
a focal area of research and training. A variety of laboi-atory and 
test locations are very Involved in space-related research and test 
activity to examine ^.he conditions, hazards, and potentials for 
pursuing missions In the space environment. 

The Space Technology Center at Kirtland AFB is a relatively new 
organization established to support the space mission of the Command. 
Subordinate to the Space Division, the Space Technology Center supervises 
the Geophysics, Rocket Propulsion, and Weapons Laboratories. Planning, 
development and test activities support a wide range of space ventures, 
with the Strategic Defense Initiative research project : prime client. 

Transforming the resource and organizational capability of AFSC 
into satisfaction of aerospace missions and objectives means the blending 
of Command activities into orchestrated efforts to meet challenges. 
Actions to pursue Command challenges in a productive and cost-effective 
manner will be discussed in the context of people, resource, and program 


One vital segment of the productivity program of AFSC rests in 
people programs. In August, 1984, General Lawrence Skantze assumed 
command of AFSC and indicated a concern for productivity initiatives 
within the Command. While entrusted with thousands of talented people 
and a large fiscal and acquisition responsibility by the Air Force, 
General Skantze saw that executing the mission of the Command would 
soon involve "doing more with less." The vast array of program 
responsibilities with increasing costs would test the ability of the 
Command to respond. 

Changing the Cultur e 

Contemporary management research and writing attests to the need 
for assessing the culture of organizations to verify goals and values 
in pursuing goals of excellence, change, or renewal. The culture change 
process is at work in AFSC. As is often the case in a change of command, 
the new commander will establish his style and working expectations in 
carrying the organization to the accomplishment of missions. Since AFSC 
is a linking pin between the using operational commands of the Air Force 
and the community of research and development (R&D), acquisitlor;, and 
civilian contracting firms, the Command finds its«>lf a military organi- 
zation with multiple ties to segments of the civilian world. 





V .». i.if>^,k vK 



General Skantze [3) has placed emphasis on managing the Air 
Force acquisition role in a "business-like" manner. This means that 
AFSC military and civilian personnel not only must meet military 
standards, but must be able to perform creditably in dealings and 
negotiations with civilian contracting organizations and R&D enterprises. 
AFSC personnel performing well wij.l best support their colleagues in 
the operational and combat conirands. "Buying smartly" would become more 
than a catchy slogan. 

Managing intelligeni.-ly includes a deep concern for quality. 
Quality assurance is a major element of AFSC acquisition and contract 
management programs. Quality in staff work and business management 
programs of the command are endorsed as a corollary of the desired 
quality sought in procurement activities. 

People Programs 

An array of ventures to Involve Command personnel in improving 
productivity and quality operations Includes: 

1. Suggestion Programs which have been emphasized to 
generate Ideas submitted to propose improvements. The 
quantity and quality of suggestions has been gratifying 

In response to Command emphasis. While a common management 
tool, suggestion systems have reflected commitment of 
organization member':. 

2. Sensing sessions with senior officers from field 
commands have assessed problems and opportunities in 
those organizations, as well as examined inter-organlza- 
tlon relationships. Input from those key personnel has 
been Instrumental in working out kinks in operations, and 
in reducing we-they confrontations. This was one of many 
innovations with behavioral science flavor. 

3. Off-site sessions by general officers and key staff '• ^ 
reviewed objectives, values, weaknesses, strengths, '*< 
threats, and opportunities which face the Command. These ' 
work meetings (in civilian clothes) promoted creative and 

cohesive options for future Command operations. ' 

4. An organizational survey of the AFSC headquarters was 
conducted by the Air Force Leadership and Management 
Development Center to provide data to managers of 
organizaclonal components. These data cover job satlsfac- 
tlor, attitudes toward work group and supervisors, organi- 
zational communications and other topics. These findings 
should provide opportunities for improving management and 
productivity in the headquarters. 




5. A civilian personnel work, force effectiveness study 
group has been convened with representation from all field 
units. This grcap is generating initiatives for optimizing 
the contributions of the civilian work force, an important 
segment of the AFSC family. 

6. Enhanced career development programs are beinf put in 
place to provide career pathing for personnel. Career 
satisfaction and progr&ss is important for the support of 
individual and work group productivity, as well as to 
provide work force stability. 

7. Decentralization is encouraged, whenever possible, as 
a means of establishing the importance of the "work place" 
and the "work unit." Senior officers, such as General 
Larry Welch, vice chief of staff (5), h: . <■ endorsp-i 
decentralization as a move to promote part- lipat ion of 
personnel at all levels in pursuing quality and proii-ic- 
tivity objectives. 

The foregoing are some examples of an emphasis on people programs 
in AFSC which are designed to promote ah awarene .; of quality and pro- 
ductivity in the culture of the Command. The entir ' Jommand has beer 
sensitized to the need ior greater attention to the use of people and 
other resources to meet goals of each organization in AFSC. Working 
"smarter" and "hetter" is encouraged as a means of dealing with the 
continually increasing workload of the Command. 


Since AFSC is entrusted with a major segment of the Air Force 
buying role, the use of sophisticated and contemporary tools and 
techniques are required to manage the development, test, and acquisition 
activities of the Command. Some of the major productivity impacts 

Management Information Systems 

To administer the functions iuherent in the AFSC nission, the 
management of information is essential. Vast amounts of data must be 
selectively acquired, stored, analyzed, and presented to decision makers 
to manage the Command. Executive Information System (EIS) capability 
is available within and between elements of the command to maintain 
real-time communications. While electronic mail and tiling (among other 
j electronic capah 'lities) are common place in industry, the military 

4 necessities, of security, readiness, and global coverage ac:entuate the 

EIS contrib-ition. 


- - '*ii~^- 


A new staff directorate (KR) was formed in 1984 to deal with 
computer resources and management information systems. The system 
architecture is being developed, with growth and versatility of function 
a key element. This capability includes use and networks of micro- 
computers, as well as networking with main-frame and support computing 
facilities. It is worth noting that AFSC facilities include equipment 
ranging from p»inl-computers to super-computers. Needless to say, 
extensive design and training activity is Involved to achieve and main- 
tain the productivity that can be attained from electronic capabilities. 

Industrial Modernization Incentives Program 

Typical of many productivity ventures in AFSC is the Industrial 
Modernization Incentives Program (IMIP) • IMIP, simply stated, is 
contracting for productivity. IMIP uses traditional and innovative 
contracting techniques to solicit, "incentlvize" and sustain contractors 
ir Increasing productivity. 

As used, IMIP is a partnership between a contractor and the 
Air Force directed at systerutically bringing the latest manufacturing 
technologies, and the capital investments needed to prodi^e them, onto 
the production floor of a contractor's facility. Resulting efficiencies 
in production yield savings to DOD weapon system contracts which are then 
shared with the contractor. This allows the contractor to realize a 
satisfactory return on the investment. 

AFSC has 32 IMIP's on contract, which inclu'es both prime and 
subtler contractor." within the aerospace industrial base. The Air 
Force's commitT.tint of over $400 million for factory analysis and 
manufacturing technology development has been matched by a $1.4 billion 
contractor ^.ommltment for development of new manufacturing technology 
and acquisition of capital equipment. The projected savings as a result 
of these commitments exceed $4.5 billion on DOD production programs 
through 1990. 

The use of IMIP combines the best of technology and good 
manufacturing management to enhance production capability. In high 
cost, short production run, and long lead time situations as found with 
space system projects, the IMIP venture is showing great productivity 

Productivity is a challenging objective in advanced technology 
programs, with costing a particular nemesis. The "Should Cost" programs 
attempt to determine ''hau costs reasonably ought to be rather than costs 
are or have been. Th'^re is evidence that the cost curve can be bent, 
rather than continuing to escalate [1], IMIP is then seen as a tool to 
help combat the serious t.ireat to productivity by program instability 
and cost overruns. 



■»*^ ^\X ,' .>^. 


Better management for productivity involves improved planning 
with accompanying rcutrol systems. In the quest for more productivity 
in its mission areas, AFSC uses planning extensively. A prime example 
is the Space Technology Plan (4), a product of the Spc,:e Technology 
Center. This type planning has a strong impact on the Command and 
relationships DOD, NASA, other government agencies and industry. 

Space technology planning focuses all Air Force space technology 
investment and execution in support of future space mission requirements. 
Selected space technology development projects will pursue objectives 
in: on-board processing and hardened electronics, spacecraft autonomy, 
space power, and advanced military spacecraft capability. 

Artificial intelligence, power cell research, and advanced 
space computer technology are hoped to provide increased spacecraft 
lifetime, memory, sun.'ivability and autonomy. The productivity of 
this research is tied to unique features of space rieeds. Some of those 
factors include: 

1. Space systems have lotig lead times and long life times. 

2. Technology development must start very early before 
systems are well defined. 

3. There is a premium on accurate forecasting of 
technology needs. 

The complexities of technologies of development, test, and 
production make the tasks of AFSC challenging. Partnerships are being 
forged with other agencies and industry to prov^ide proactive approaches 
to managing for quality outcomes in a cost-effective manner. 


A recent NASA report on improving Quality and Productivity in 
Government and Industry [2] captures the essence of the AFSC quest for 
productivity in its activities. Quality of work, quality of work life, 
and quality of management are cornerstcnes for increasing productivity. 
Technology and production bases help provide the setting for quality 
of work in meeting standards and requirements for products and outputs. 

People programs generate commitment and work environment 
conditions which contribute to the success of th^ organization. Quality 
of management fosters leadership that has the technical and leadership 
skills to provide direction and feedback in guiding the organization 
to its objectives. The people of AFSC are working toward those quality 
conditions in the pursuit of productivity. 




(1) Correll, John T., "Bending the Technology-Cost Curve," Air Force 
Magazine , Vol. 66, No. 8 (August, 1983), pp. 45-48. 

{2] NASA, A Framework for Action; Improving Quality and Productivity 
in Government and Industry , December, 1984. 

[3] Skantze, Lawrence. Personal Communication. July, 1985. 

[4] Space Technology Plan , Space Technology Center, Kirtland AFB, NM, 

[5] Welch, Larry D. Management and Productivity . Remarks to the 
National Forum of Human Resource Planning, Baltimore, MD. 
May 9, 1985. 

Major General Vince Luchsinger is a member of the Air Force Reserve and 
serves as Mobilization Assistant to General Lawrence Skantze, commander 
of the Air Force Systems Command at Andrews AFB, MD. He also serves 
as Dean of the School of Business at the University of Baltimore, 
Baltimore, MD. 






D. H. Hutchinson 

McDonnell Douglas Astronautics Company 

Huntington Beach, California 


Eighty-three management Improvenent pro.iects Initiated f*^ the 
McDonnell Douglas Astronautics Company since April 1983 provide case 
studies of productivity Improvement techniques; *n& measurenenis of 
improvenent . These projects were Initiated as part of a program of 
self-renewal that Sanford N. McDonnell, McDonnell Douglas Corporation 
chairman and chief executive officer, outlined In his Mai 1984 message 
to the corporation. He described the following five initiatives 
designed "to foster self-renewal throughout the corporation and thereby 
prepare for the future": 

!• Strategic Management . A continual dynamic open-ended 
planning process that focuses on overall goals, long-range needs, 
problems, and opportunities to ensure that the corporation grows and 
meets the requirements of an ever-evolving marketplace. 

2. Human Resource Management . The company assures Itself of a 
continuing cadre of top-flight peoiOle by helping employees plan and meet 
ceweer goals and by identifying and grooming those with the greatest 

3. Participative Management . A two-way process that allows 
employees to share in the shaping of the future of the company by 
encouraging people to express their opinions and by training managers to 
listen to those opinions and consider them very carefully. 

4. Productivity Improvement . Productivity: another word for 
efficiency — for achieving the maximum In product value while avoiding 
unnecessary costs both in house and at vendors to meet the challenge of 
aggressive competitorss. 

5. Ethical Decision-Making . Adherence to high ethical 
standards is a commitment to excellence and a central part of the 
Corporation's traditions and world-wide reputation. 

This paper discusses productivity improvement: Sandy McDonnell s 
fourth key for corporate self-renewal. The concept of productxvlty or 
efficiency is supplemented with two additional concepts of productivity 
improvement: effectiveness and innovation. Dr. V. Edwards Demlng, 


^•»"f4. c^-- '^\ ' ■ ' ^ ' ('^ 

early In the 19508, applied statistical techniquep to the Improvement of 
quality [2], and Dr. Kaoru Ishikawa worked with the Quality Circle 
Headquarters of the Union of Japanese Scientists and Engineers in Tokyo 
[4]. These activities provide the basis for the well-documented process 
of improvement of efficiency and productivity in several American and 
Japanese industries, and for the process of continuous improvement in 
the McDonnell Douglas Corporation. Case studies of improvement of 
efficiency, effectiveness, and innovation at McDonnell Douglas are 
presented in this paper. 


Strategic objectives and business unit planning are the framework 
for quality and productivity improvement at McDonnell Douglas 
Astronautics Company in Huntington Beach, California (MDAC-HB). 
Business units, organized ai I'und products, are expected to manage 
continuous improvement during product life cycles. The mission, focus, 
and methods of improvement are adapted to the achievement of program 
goals during product life cycies. The methods must afford continuous 
improvement. Mission and focuo are adaptive as the program life cycle 
moves from conception through design and development to operations a.nd 


The dileimia facing American industry has been well documented. 
In industry after industry, US firms are losing their competitive 
advantage. This loss is the underlying cause for the country's present 
economic distress and many relateu social problems. In an effort to 
turn this loss around, industrial leaders have sought ways to renew the 
competitive edge. The readaptive processes of the automobile, steel, 
hospital, agricultural, residential construction, coal, and 
telecommunications industries have been well described [7]. The 
aerospace industry is not exempt from world-wide competition; It also 
must renew itself. The readaptiva process in the aerospace Industry has 
begun to receive the combined attention of government. Industrial, amd 
educational institutions. The approach used at McDonnell Douglas 
includes performance improvement project teams (Appendix) that focus 
upon and solve problems as part of a continuous process of Improvement 
(Figure 1). 

The forces that lie behind the form that readaptatlon takes In 

the seven basic industries studied by Lawrence and Dyer [7] and those 

giving form to the readaptive process at McDonnell Douglas are similar. 

• The most remarkable are changes In core technologies, policies, 

competition, and the Importance of choices In organizational strategy. 


> The readaptive process (Figure 2) Involves seeing organizations 

^ as soclo-technlcal-learnlng systems. The Idea that orgarlzatlons must 








Focus on 





Yield of Resources 

Focus on 





Focus on 






Definition and 

Design Phases 

Design and 






Intermediate Resource 

intermediate Information 



■ Readaptive Form 

• High Differentiation 

• High integration 

• Balanced Clan, Market, and Bureaucratic 
- Human Resource Practices 

• Power Balance, Vertical and Horizontal 

■ Readaptive Strategy 

involvement of Memtiers 
in Learning and Striving 




. -.-J*.., 


iKN-;'i:.L^ «.'•:. ■- 


be learning syetems is not new; it was the central theme of James March 
and Herbert Simon's classic 1958 work [8]. It is also the basis of 
rigorous studies of adaptive systems technology [9] . What is involved 
is a change in ways of looking at organizations. The old concepts of 
mechanical efficiency and bureaucracy are replaced. An integration of 
innovation and efficiency must take place through the involvement of all 
of the employees in active processes of learning and striving (effective 
interaction). At each stage in a program's life cycle, the process of 
performance improvement must develop an appropriate focus so as to 
improve quality and productivity. The focus of effort will change as 
the task evolves. Early phases of programs are chtiracterized by the 
need for purposeful innovation. 

Purposeful Innovation ( Doing Well What Has Never Been Done Before) 

The early stages of a program life cycle focus on purposeful 

innovation. ~ novation la the heart of renewal. C;ise studies from 

S Renewing American Industry [7] suggest that if organizntions are to be 

I efficient operational systems, they must also be Innovative learning 

"j systems. Systematic innovation consists of the purposeful and organizea 

J- search for change, and the systematic analysis of the opportunities such 

; change might offer [3]. The term innovation as U3ed here ia an 

1 economic, social, and technical terqi. It is the improved yjald of 


E ffective Interaction (Doing the Right Thing — The Integration of 
■; Innovation and Efficiency) 

Organizations are interacting sytems in the process of 
T development. New programs demsmd renewal of theae systems. 


1 As programs reach the development stage, each department must 

7 learn to adapt and effectively Interact with other departments. 

Effective interaction occurs when producers know what users need and 
produce what satisfies them. Successful programs depend upon effective 
Interaction. Improvement projects must focus on producing useful 
information: budgets, schedules, contracts, tests, software, and so 
forth. The interactions of these services provide a source of 
'i organizational effectiveness improvement. Improvement in interactive 

; effectiveness Is prerequisite to Improvement In operational efficiency. 

* One must be doing the right thing, then improve the efficiency of the 

operation. Effectiveness Improvement is a task of the developing 

Operational Efficienc y (Doing It Right the First Time) 

As programs move toward operational phases, the focus shifts 
again, this time toward process control. The central focus of the 
productivity movement embraced in Japan after World War II was 
operational efficiency. The work objective was to do it right the first 
time. Doing the right Job correctly is the key to productivity. Through 
the development of quality circles in Japan, this resulted in the 
Japanese worker's having almost total authority within his discipline. 



•-. 'i&^-Tr- 

.N**'^''^'--^ '' ■ H\ 

There Is a convergence of interests that amounts to participative 
management . 

Many companies in Japan and in the United States are currently 
involved in Intensive efforts to incorporate quality concerns into 
problems of product-line trf-^sitions, applications to design, relations 
with customers, and research and development. Although operational 
efficiency is necessary in these functions, it is not suff^.cient. 
Effectiveness amd innovation are also critical. 


i Operational reliability, quality, and produr-lvlty depend upon 

^ having all of the parts working right in a process. When organizations 
J like McDonnell Douglas and NASA take on new projects, they take on tasks 

of organizational transition. Treinsitional adaptation through the r-lfs 
cycle of a program, from concept development to operationally successful 
missions, includes an adaptation of performance improvement techniques 
to the changing nature of the problems encountered. J. M. Juran has 
called improvement projects "probl^ns scheduled for solution" [5]. The 
problem-solving methods pioneered by Dr. W. Edwards Deming [2], Dr. 
Kaoru Ishikawa [4], and Dr. Juran in Japan have resulted in the 
continuing development of a management improvement process . This 
process is an ideal vehicle for continuing to improve quality and 
productivity as a program makes the treuisition from research and 
development to operational maturity. 

The current Space Station program, committed to putting a 
permanent manned station in space by the early 1990s, is in its 
definition and preliminary design phase. Thus, Space station program 
performance improvement projects must be devoted to purposeful 
innovations. The spirit of entrepreneurship, of newness, of finding 
opportunities to do economically, socially, and technically what has not 
been done before facilitates this stage. Perspective must be shifted 
from the operational worker as a follower of practices on some previous 
project to a knowledgeable worker guided by a renewed sense of purpose 
and a sense of the new whole of which he is a part. 

Purposefully Innovative Systems (Doing Well What Has Never Been Done 

Case A focused factory or "plant within a plant" was 

Study developed at MDAC in 1979 to answer the need for 
No. 1: high-technology electronics manufacturing within 
E/E Develop- the company. The need for complex electrical and 
ment Versus electronic (E/E) products required a highly 
Production technical manufacturing capability and recent 
contract demands for concurrency of design and 
development required a quick-response capability. 
The systems of the new "plant within a plant" were 
not fulfilling this quick-response requirement. A 
reevaluation of the plant's developmental and 
production roles and missions was in order. 


*-- ^^.jr^: 

4' A'*rt'%^'" 


A Juran project (see Appendi\) was implemented to explore ways 
for the "focused factory" to meat quick- respxsnse requirements and 
expectations. A team from manage^aent and engineering %ras selected and 
tha problem was stated: existing systems were not fulfilling the need 
for the highly technical, -juick-response memufacturing capability 
required by complex pro^iucts and concurrent development of 
electrical/electronic items. The improvement project team thus set as 
its objectives to (1) ar>sess existing methods and procedures and 
structure new methods an<'l procedures aligned more in keeping with a 
developmental approach, (2) establish task definitions for the 
implementation of a developmental shop, and (3) reconoiend organizational 
and support function c.iemges . The team met periodically for nearly a 

The problem with development projects had been evidenced in 
schedule slips, erst overruns, drawing release delays, and errors in 
work instructions. Systems being used stressed control rather than 
innovation, and fitted mature products rather than development 
projects. This pattern appeared in drawing preparation, production 
planning, manufacturing, inspection, and test. Also, the knowledge 
gained during development was often lost; Job classifications were not 
appropriate, drawing systems needed changing, and organizational 
reporting relationships needed adaptation. 

The statistical (variance reduction) techniques of the Juran 
process were not very useful for this kind of performance improvement 
project. This project required a form of performance improvement that 
Peter Orucker has recently described as "purposeful innovation" [3]. 
The key recoeriiendation v«s to form a development team that has the 
flexibility to depart from the past and Invent new ways of working so as 
to adapt to the variables of developmental projects. The improvement 
project team worked out remedies In the drawing system, provided 
development work instructions, and focused responsibility for 
development tasks. More involvement in goal setting and planning was 
needed at the interface of engineering and manufacturing to increase 
flexibility and efficiency, match people to tasks, and account for 
costs. The problems, causes, and remedies are shown in Figure 3. 

Figure 3. case study no. i-e/e development versus production | 




■ Unraallttic tchadul^^ 

■ High coat 

■ Delays In drawing and production 

■ Kay contributions missing 

■ Incomplata dsvalopment holding 
up production 

I Re<^ulrement8 geared to mature 

I Work instructions difficult to 

I Departmcints geared to stable 

I Policy and practices emphs8i;:e 
control instead of flexibility 

I Development knowledge not 

• Implement development drawings 

■ Provide development work 

■ Focus responsibility for 
development tasks 

■ Encourage capture of good ideas 

■ Allow all Involved to make inputs 

■ Involve others In goal-setting 

■ Ta':or to needs of situation 


' #(<-r;. •, 




Interactive Effectiveness (Doing tha Right T hing) 

Case A pilot "white-collar" productivity lesproveoient 

Study project (WCPI) wan Initiated in MDAC's Department 
No. 2 147, Financial Controlc— Direct Budgets, in 1984. 
It has been a unique project. It addresses the 
renewal of a department. A project team of r^lne 
represented the 33 members of the department. 
This team v/as confronted with the demands of 
interactive effectiveness. 

A pilot "white collar" productivity improvement project (WCPI) 
was initiated at MDAC to explore productivity in a department whose 
function is to provide useful services and Information. The American 
Productivity Center's) six-phase WCPI framework" provided a methodology. 
This particular WCPI project Involved three groups of people: 

• The P roducers . Meetings were neld in which all 33 members 
of Department 147 worked together to diagnose problems and opportunities 
for improvement. Complete feedback was given to each member of the 
department regarding the retults of a 200-questlon survey of attitudes, 
leadership, co;nmunicatlon, participation, goals, measurement, rewards, 
resoorces, and related concerns. They then set out to make improvements. 

• The Users . One-hundred and fifty fiscal, program, and 
business managers as well as engineering and operations users in eight 
different program or business units were surveyed and Interviewed. 
Feedback from this material provided a basis for (1) redesign of the 
services of the department to neet strategic plans, (2) team building 
for new teaming alignments, and (3) technology applications for Improved 
service to users. 

• The Managers . A management group Included the Director of 
Financial Controls, the Controller, the Vice President-Fiscal, and the 
Director of Productivity. Department 147 's manager became a member cf 
the producer project teair, In an example of participative management in 

The greatest block to Interactive effectiveness in organizations 
is the attempt to protect Interest groups and preserve old ways of doing 
things. These interests must be worked through person by person. This 
WCPI pilot project provided a framework to make transition processes a 
part of personal and organizational '•enewal, a concept to be instituted 
on a department- by-department basis in selected parts of the company. 

The lessons learned, expressed by the improvement project's team 
members. Included the following: 

1. Commitment to Improvement : Producers, users, and managers 
must be committed to Improvement. 

"The American Productivity Center's six phases of a white-collar 
productivity improvement project are, briefly, diagnosis, objective 
setting, measuring, service redesign, team development, and development 
of technology parameters. 



2. People Involvement : 
nor., influences Interaction. 

Everyone's personal opinion, valid or 

3. Producer Ownership ; Each producer and each user is a stake 
holder who viust own and share in the gains of improvenent . 

4. User Involvement : Effectiveness means doing the right 
things and requires interaction between producers and users. 

5. Responsiveness : 
those who are served. 

Services must be responsive to the needs of 

6. E fficiency/Effectiveness : 
develop organizational effectiveness. 

The role of the WCPI is to 

The WCPI process works. Its improvements accrue to producers, 
users, and managers (Figure *) if they believe in it. But credibility 
must be won. In the Department 147 project, not one of the three groups 
was easy to convince. Even the pilot coordinator and managers had to go 
through the process. The key is complete honesty auid feedback. The 
first meeting with the entire department did not result in belief 
because the will to change was not apparent. It took action to show 
that manager, user, and producer groups had the will to do something. 
Once members of the department received feedback and became involved in 
diagnosis and objectlve-stting, things turned around dramatically. 
Commitment came even in the midst of the heavy daily workload. 




*0- t-^- 




The WCPI resulted In the Improvement of attitudes, participation, 
connunlcatlon, leadership, goal-setting, measurement, rewards, 
recognition, and use of resources within the department. It resulted In 
the Involvement of management not only In getting to know the work of 
the department moro thoroughly, but In providing a sense of mission and 
focus and In selecting the team and then allowing the team to actually 
make Improvements. The UCPI also resulted In Improvement of the 
relationship of the producers to managers and users (Figure 5). This 
began with Identlflcaton of the users continued with attempts to see 
the department through their eyes, and resultwd In further Improvements 
that will continue to be made In the future. 


6 Interfaces and External Forces 

5 Community Culture 
and Organization 



5 Community Culture 
and Organization 


1. Production Processes 

2. Utilization Processes 

(Adapted from Killman [6]) 


5. Community Culture and 
3. Dissemination/Provision 

Processes 6. Intarfaces and External 


Operational Efficiency (Doing Things Right the Clrsl Tiac ) 



No. 3: 

Crush Grind 

Process Control 

Hardware acceptance is determined after 
manufacture by part Inspecton. However, 
Inspection does not prevent nonconformance 
of parts. Since a significant percent of 
machine shop rejections is accounted for by 
machine operation, more hardware could be 
accepted if the manufact jring procsss could 
be control ie' so as to prevent nonconformance 
of parts. 



- *!*■■.>-.* 


or;g::". . : 


NDAC had recently been discusb^ng the aierlts of a quality control 
philosophy based upon prevention rather than detection of defects. To 
test this concept and solve this specific parts-rejection proolem, 
nanagenent chartered a team representing manufacturing, planning, a"id 
quality departments. 

Through analysis of prior nonconfomance costs, a particularly 
unsatisfactory nachlnlng process was identified. This process Is 
required to hold very close dimensional tolerances, out It consistently 
produced defective parts. Many traditional approaches to corrective 
action had been tried but had met with little success. Thus, the team 
felt that a new approach was In order. 

The team used the following Investigative procedure: (1) 
Identify likely causes, (2) test these possibilities to determine their 
actual effect, (3) evaluate test results and Inpleuent corrective 
solutions, (4) test tliese solutions for effectiveness, (5) train 
operating personnel In the new approach. (6) reduce or ellmlnat** ln-l.».ne 
Inspection and accept the product on the basis of process control. 

To Identify potential causes of defective paits, the team held a 
brainstorming session. Over 75 opinions or possible causes were 
gathered through this process . These opinions were Cv ^egorlzed 
diagraflBied on an Ishlkawa "fishbone" chart [4] to show their 
interrelationships (Figure 6). Next, copies of the fishbone chart were 
distributed to persons familiar with the machine procecs under 
investigation. These people were asked to identify tha ten most likely 
causes of process nonconformance. These results were then correlated to 
produce a list of the most likely cnusec. 

The group then determined what data and experimental approaches 
were required to prove or disprove these most likely causes. In this 
case, a proc ss capability analysis was considered crucial to evaluate 
the possible defect causes. 




PERSO^»■CL suPPtlE^ ^l^^ll^ 



irnvM* IBH » li^<«l 

I MMM U*i*% ■/ 





• • '^t^p -. '' 



^P>S-"^^"-. ■ 


Once the experiment was performed, the data were analyzed with 
simple statistical techniques (Figure 7). These statistical tools 
Included histograms (Figure 8) that showed how the data were distributed. 



1.6180 - 


5 1.6176 



i 1.6172 




S 1.6168 


1.6164 - 


■ = Nonconformance 

Specification ■ ■ 
Limits ■ 

!„„„■-. ■„ „ 

|- rm " ^ mrrm nn r^ r^ d 
1 d'"^„„ „.„ □ tn u 'iiLu u nil iiiii d or 

m n rm 
^ uju Li^ on D D c^a nn 03 

a a □ D 




^ ■ ■ 



_ L _ 1 . 1 ._ 1 1 1 1 1 1 


10 20 30 40 50 60 

Parts in Run Order 




Diameter (1.6160 to 1.6185) 




The andlysis revealed several significant conclusions. Of 
greatest iicportance, it revealed that the process was not capable of 
consistently producing parts within the design tolerances. In fact, the 
experimental results suggested that a 25% failure rate should be 
expected fron this pro^sess. Differences in inspection techniques 
between operators and inspectors used as such as 20% of the design 
tolerance. Variations in part center depths <this process grinds parts 
on centers) were as high as + 0.029 in., requiring a new setup for each 

Clearly, this was new and interesting information. Here was a 
machine process destined to make defective parts regardless of the 
efforts of the operator. To correct this obviously unsatisfactory 
situation, equipment was ordered that could be attached to the machine 
to improve its capability and automatically compensate for variations in 
part center depth. This equipment cost less then $20,000 and will pay 
for itself in less than 1-1/2 years. In addition, manufacturing and 
inspection agreed to use the sane inspection tools and techniques. 

Once the new equipment is received and installed, the study team 
will perform a new process capability study to prove that the machine 
enhancements are effective. Once this is demonstrated, the machine 
operators will be trained in simple statistical techniques, especially 
control charts. The operator will then implement normalized process 
control cnarts and Quality Assurance will inspect by monitoring the 
control charts and auditing the manufacturing inspections. 

This project has been very successful so far and has served to 
show the viability of the process control approach to quality control 
[10]. In addition, it strongly suggests that a process isn't thoroughly 
known until it is known statistically. 

KDAC anticipates more projects of this kind in the future, 
leading to significemt improvements in quality and reductions in costs 
through improvement in operational efficiency, in office technology as 
well as in manufacturing. 


The mission, focus, and methods of quality and productivity 
improvement projects are part of & renewal or readaptation process that 
changes during the life cycle of aerospace programs. The industry is 
beginning to become aware of the need for managing life-cycle 
transitions and developing self-renewing organizations. Uilllam 
P'^dges, a consultant to HDAC-HB, has summarized the challenge of 
organizational transition [1]: "'Change' is when something stops 
happening or when something starts happening. 'Transition,' on the 
other hemd, is an inner reorientation process that an organization and 
all the individuals in it go through when a change requires everyone to 
work on a new team, fill a new role, and act in new way." 



^, > r~!.''%v'-. '- 




Juran Piolact Taenia 

1. TMm eomiiosltlon/ 

4 to 10 mambara on an 
IntradMalonal or 

managar, coordinator, and 
APC Iblaon 

S to IS mantbara, not 
nacaaaarily from tha 
aama work group, arith a 

L PwtleipMlor> 

Not voluntary; natuta of 
prolacta aalactad 
datarmlnaa participation 

cf valopad In tlia procaaa 

VolunUry lor circio 

3. Pra|«ct Mlactlon 

Prolacta aalactad by 
Improvamant Pro|act 
(MIR) Council 

Prolact aalactad by MIP 

Prolacta aalactad by 
taam mambara 

Unw raqulrad 

Complax pfOtNama 
raqulring aubatantlal 
tima (or analyala 

Foeua on organizational 

Modarataly complax 
proWama raqulring lima 
tor analyala 

S. Training 

Entira taam tralnad In 
problam solving and 
autlatlca; laadar training 
lor taam laadar 

Training provMad by tha 
APC) and proiact coordinator 

Entira taam tralnad In 
axtav-itlva laadar training 

6. Facilitator or Intamal 

FaclllUtora and 
atallatlclans avallabia 

raqulrad; llalaon aupport and 
contoratKaa at APC 

Facllltatorw and 
atatlatlclana avallabia 

7. Coordination 

MIP Coordinating 
Commlttaa ar ' 
Functional Stearing 

Pilot prolact coordinator from 
tha company aaalgnad to work 

Coordination by 

S. DMlakm-maklng 1 HIglwr authority 
iiuthorlty requlrad to declda on 
j recomnwndad actions 

Rallas on partlclpaUva 


LImHad authority to 
daclda on action; maka 
racommandattona to 



tl] Bridges, WllliaiB, Managing Organizational Transitions 
(unpublished paper); Transitions: Making Sense of Life's Changes . 
Addison Wesley, 1980. 

[2] Demlng, W. Edwards, Quality. Productivity, and Competitive 
Position . Massachusetts Institute of Technology, Cambridge, 1982. 

[3] Drucker, Peter F., I nnovation and Entrepreneurship . Harper and 
Row, 1985. 

[4] Ishlkawa, Kaoru, G uide to Quality Control . Asiem Productivity 
Organization, Tokyo, 1974. 

[5.1 Juran, J. M. , Managerial Breakthrough . McGraw-Hill Book Company, 
New York, 1964. 

[6] Klllman, Ralph H. , et al.. Producing Useful Knowledge in 
Organizations . Praeger, New York, 1983. 

[7] Lawrence, Paul R., and Davis Dyer, Renewing American Industry . 
New York, Macmillan, 1983. 

[8] March, James, and Herbert Simon, Organizations . New York, Wiley, 


[9] Tsypkln, T. Z., Adaptation and Leat-nln£ In Automated Systems . New 
York, Academic Press, 1971. 

[10] Toung, John A., The Quality-Productivity Connection . American 
Electronics Association Conference, November 8, 1983. 


D. H. Hutchinson is currently a coordinator and Internal 
consultant to leaders of mejiagement Improvement projects at the 
HcOonnell Douglas Astronautics Company. He has been a facilitator and 
trainer on the Initiation of Juran projects and the white collar 
productivity improvement pilot project as well as providing support to 
training and connunicatlons in the "Five Keys to Renewal" since the 
inception of these initiatives at HDAC Huntington Beach. Hutchinson has 
gained extensive experience in Engineering, Program, Human Resources, 
and Operations departments during his 29 years at McDonnell Douglas. He 
received his BA and M.Ed degrees from UCLA and is currently in the 
doctoral program at Claremont Graduate School. His dissertation, 
entitled "Resymbolization of Work", deals with a theory of 
knowledge-tfork productivity and will b« published in 1986. 






I : 



V^S^""*" . 

W ^ ■ -.-iMl cB't. 




William A. Johnson, Rockwell International 
Christian T. Thomasen, Rockwell International 


STS operations managers are being confronted with a unique set of 
challenges as a result of increasing flight rates, the demand for flight 
manifest/production schedule flexibility and an emphasis on continued 
cost reduction. These challenges have created the need for an integrated 
production planning system that provides managers with the capability to 
plan, schedule, status and account for an orderly flow of products and 
services across a large, multi -discipline organization. With increased 
visibility into the end-to-end production flow for individual and 
parallel missions in process, managers can assess the integrated impact 
of changes, identify and measure the interrelationships of resource, 
schedule, and technical performance requirements and prioritize produc- 
tivity enhancements. 


The Operations Challenge 

The Space Transportation System (STS) flight-to-flight reconfig- 
uration process at Johnson Space Center (JSC) encompasses initial flight 
feasibility assessments, flight trajectory design, flight operations 
planning, flight production generation, facility reconfiguration, flight 
I crew and controller training, real-time mission support and post-flight 

reconstruction. As the STS transitions to mature operations, this 
end-to-end process must be streamlined to minimize cost while 
maintaining acceptable quality and risk levels. Reconfiguration 
schedules must be reduced while maintaining late manifest and flight 
content change flexibility. Increased flight rates must be accommoc'ated 
within production capacity and manpower limitations. These challenges 
have created the need for an integrated project/performance management 






system to provide managers with improved visibility and control of the 
reconfiguration process flow and the flight production line. 

Production Line Environment 

Each STS mission flown increases the experience database and the 
potential for a higher degree of commonality between a new set of 
mission requirements and those of a previous flight . Continued stan- 
dardization of flight products and configurations promote maximum 
utilization of generic data and minimizes th effort required to generate 
flight specific updates. Incorporation of other process enhancements 
such as such automation, modularization, consolidation, simplification 
and elimination contributes to a changing operational environment that 
parallels a production line methodology. As management focuses on the 
challenge of maintaining an orderly flow of products and services across 
the large , multi-discipline organization responsible for the JSC 
flight-to-flight reconfiguration process, the supporting 
project/performance management system must be oriented to provide the 
capabilities of an integrated production planning system. 

Reconfiguration Process 

The flight-to-flight reconfiguration production process is 
composed of three major elements. First, the activities, products, 
schedules and associated resources within each technical discipline that 
are structured to meet flight specific requirements and milestones. 
Second, the interrelationship of the flight preparation, production, 
mission operations and support functions that establishes process 
constraints between technical disciplines for a specific flight. Third, 
the composite set of flights in serial and parallel processing that 
compete for utilization of the common manpower, facilities, hardware and 
software resources. The end-to-end process covers approximately a one 
year period from the Flight Definition and Requirements Directive (FDRD) 
baseline through launch of the mission. It is a relatively dynamic 
process with the possibility of changes in the flight manifest config- 
uration, vehicle assignments, flight schedules or mission requirements 
occurring late in the flow. The impact of these changes may be to a 
single mission or domino into several flights. 




Integrated Production Planning System 

The integrated production planning system must provide managers 
with the capability to plan, schedule, status and account for all the 
work involved on a single and multimission basis within this dynamic 
environment. As a resource management tool, the system must support the 
integrated forecasting and performance assessment of the resource 
requirements and resource-leveling techniques where necessary to stay 
within production capacity limitations. As a work process analysis tool, 
the system must facilitate the identification of schedule conflicts and 
production process choke-points and support the evaluation of productiv- 
ity improvement options and determination of relative priority for 
implementation. As a change assessment tool, the system must provide 
data which clearly defines the integrated impact of the management 
decisions to be made on a single mission basis and the potential domino 
effect to parallel in-process missions. As a model of the flight-to- 
flight reconfiguration process the integrated production planning system 
can provide an invaluable tool in the successful management of STS 


Activity Categorization 

The methodology used in developing a model of the flight-to-flight 
reconfiguration process is based on a building block approach from an 
individual activity level to an integrated flight-specific work plan, 
and ultimately, into a composite multimission work plan. Initial analy- 
sis of the NASA and contractor work requirements identified in three 
basic activity categories: 

1) Manifest-dependent activities that are dedicated to or directly 
associated with specific flight schedules and requi. ements. These 
activities (e.g., the Orbiter on-board mass memory unit (MMU) load 
production) are generic for each mission. 

2) Manifest-related activities that are institutional in nature, 
but with periods of dedicated support to a specific flight. These 
activities may be derived from the composite multimission resource- 
utilization profiles associated with the manifest-dependent activities 


5) - ^ 

"^23^ ' 

(e.g.. Central Computing Facility (CCF) operations) or added to the 
generic manifest-dependent activities for a specific flight requirement 
(e.g.. Shuttle Flight Operations Manual (SFOM) update). 

3) Manifest-independent activities that are regularly scheduled, 
on-going or discrete project activities having little interaction with 
flight requirements. These activities (e.g., CCF preventative mainte- 
nance) would be integrated with the manifest-related or independently 
scheduled (e.g.. Integrated Management Center (MTC) development). 

Generic Process Flows 

The manifest-dependent activities with their associated resource 
requirements (manpower/facility/hardware/software) are the basic stan- 
dard building blocks for the model. Generic process flows can be con- 
structed for each technical discipline linking the activities with their 
key input and output products, milestone events, and serial or parallel 
time-phasing for a typical high complexity mission. Integration of these 
generic process flows for each technical discipline into an end-to-end 
network logic flow establishes the interrelationships and process 
constraints between disciplines. The data can be structured into several 
levels of hierarchy for corresponding levels of management visibility. 
Hammock activities are established at higher levels to span the compos- 
ite duration and summarize the status of a group of lower tier activ- 

Key milestones can be fixed at a given launch-minus date to 
conform to top-level tempU^-es previously negotiated. The number of 
these fixed milestones should be kept to a minimum, however, this allows 
activities to float between the earliest possible start and the latest 
possible finish dates imposed by input/output dependencies with other 
portions of the network process flow. As discussed later, this factor 
becomes critical in the ability to level the utilization of resources. 

Flight-Specific Tailoring 

Flight-specific tailoring is required to transform the generic 
network logic flow into a series of flight-specific plans consistent 
with each set of mission requirements. An analysis of the relative 
mission-unique drivers and complexity factors war performed to determine 
how one mission differs from another. The first conclusion was that M've 
levels of complexity are desirable to tailor activities, products, 
schedules, and resources to mission specific requirements: high, nedium. 






low, DoO, and launch-on-need (LON). Second, different disciplines are 
affected by different mission-unique drivers and complexity factors: a 
single overall rating for the total network logic flow Is not always 
applicable. Third, complexity can vary as a function of the mission 
phase. Fourth, an approach that considered relative complexity by 
mission phase and discipline provides an opportunity to reduce selected 
portions of the network logic flow where a uniform complexity reduction 
of the total flow is not possible. Development of a parametric matrix 
approach to the mission-unique drivers for each mission phase allows a 
judgement weighting factor of relative complexity to be applied to a set 
of specific flight requirements. 

F light-Specific Plans 

Application of the mission-unique driver matrix to a projected 
flight rate model and generic network logic flow results in a series of 
flight-specific plans. Activity durations and the associated resource 
requirements for lower complexity tasks can be reduced accordingly. 
Adjustments can be made for utilization of dedicated DoD secure re- 
sources where applicable and manifest-related activities added where 
appropriate. While each flight-specific plan conforms to the same 
top-level template of key milestone launch-minus dates, flight-specific 
tailoring by technical discipline and mission phase results is essen- 
tially a unique critical path for each plan. Each flight-specific plan 
is also tailored to achieve the minimum production cost by consideration 
of the relative mission complexity factors in developing the activity 
duration and associated resource requirements. These schedule and cost 
targets can be planned a ' asured on a flight-by-f light basis. 

Multimission Work Plan 

A composite multimission work plan is created from the individual 
flight-specific plans. Each activity and event is coded to identify the 
associated flight number, responsible technical discipline and organ- 
ization, authorizing work breakdown structure (WBS) code and other 
selection/sort flags to facilitate a variety of composite output reports 
and graphics. Other manifest-related and independent activities can be 
overlayed, to complete the scheduled work profile for any given time 

Analysis of the multimission work plan data can identify potential 
schedule conflicts, production flow choke-points and timeframes where 
the resource utilization requirements exceed the composite production 




V'ttSKx-* . 

^< .i?;i''*«fe, :^<h^.■ 



capacity and manpower limitations. Automated resource-leveling can be 
performed by establishing a resource availability curve and allowing the 
computer to reschedule the effected activities to a more opportune 
timeframe within the network logic constraints for each mission. Groups 
or categories of activities can be given relative priorities to influ- 
ence the automated rescheduling process. Successful resource- leveling 
may not be possible if the network logic does not provide sufficient 
schedule flexibility with activity float between the earliest and latest 
possible start dates. 

The impact of a change in mission requirements can be assessed by 
revising the appropriate flight-specific activities, products, mile- 
stones, interrelationships and associated resource utilization require- 
ments. The revised flight-specific plan(s) can be incorporated into a 
composite multimission work plan to evaluate the integrated impact of 
the change to the current baseline or combination of baseline and other 
pending changes. This same approach can be applied to evaluation of 
proposed productivity enhancements. Revision of the generic network 
logic date corresponding to the improved process flow or revised re- 
source requirements can be projected to a multimisslon scale to deter- 
mine overall gain in production efficiencies. 


System Prototype 

A prototype of an integrated production planning system was 
developed utilizing the PROJECT/2 project management software to vali- 
date the concept and methodology described above. PROJECT/2 software was 
selected for the prototype based upon the availability of a mainframe 
installation within our corporate resources, which met the system 
performance criteria. Other considerations Included accessibility of 
experienced users, and a demonstrated user-friendliness. The key factors 
In the success of any project management system are the structuring of 
the database which must simulate as accurately as possible the 
real-world environment and the capability to present the data in formats 
which managers can readily assess plans, status, analyze alternatives 
and incorporate changes. The focus of the prototype develofjment was to 
demonstrate these capabilities. 


^1 Aj^^\I^>^^V 



A generic network logic flow was created interrelating over 500 
activities, products, and milestone events to model the JSC 
flight-to-flight reconfiguration production process. The structure of 
the network logic flow was built around the definition of 11 technical 
disciplines which provided a better validation of their functional 
interrelationships. The responsible organization, WBS coding, se- 
lection/sort flags, and associated resource utilization profiles were 
[ added to the database for each activity and event defined. Samples of 

flight-specific work plans at several levels of hierarchy were generated 
f both the graphic and tabular form. 

A composite multimlssion work plan was created combining 36 
flight-specific plans representing flights from 51-L (01/22/86 launch) 
through 81-G (02/15/88 launch). The multimission database contained over 
19,000 activities, products, and milestone events, and over 30,000 
logical Interrelationships. Samples of composite production schedules 
; and resource utilization profiles were generated. A preliminary eval- 

} uation of auto».iated resource-leveling and priority scheduling options 

ware also concluded. A data transfer approach was developed and demon- 
strated to facilitate interchange of data between the prototype system 
and the other existing project management/scheduling systems. The results 
to date have proven the feasibility of the concept and validated many of 
[ the conclusions reached in developing this methodology. 


A realistic amount of schedule flexibility (critical path float) 
must be provided within the network logic for each mission to accomno- 
date unplanned changes and resource-leveling when required. The net 
impact to a single mission and potential dcmino Impact to parallel 
missions in process must minimized to maintain overall production 
efficiency. While reduction of the end-to-end production template for a 
mission does provide a minimum schedule, it does not necessarily provide 
a corresponding reduction in the composite production costs for all 
missions if flexibility is lost. 

As the STS flight rate continues to increase, the relative time 
between missions decreases resulting in a higher degree of overlap for 
the same activities in parallel processing for different missions. This 
creates an increase in the number of equivalent missions competing for 
utilization of the same common manpower, facility, hardware and software 
resources. A shorter production template maybe needed to reduce the 
overlap and reduce the number of equivalent missions in parallel pro- 
cessing. The challenge is achieving a balance between a shorter template 
and retaining the necessary schedule flexibility. 




%x i'ii'*». Vt^i. 

One snlution is a continued emphasis on productivity enhancements. 
Standardization of flight products and configurations through uti- 
lization of generic data reduces their sensitivity to unplanned changes 
in requirements and the need for ''chedule flexibility to Incorporate the 
changes. Automation, modularization, and consolidation of activities 
generally reduce the end-to-end processing time and Increases flexibil- 
ity. Task simplitlcatlon reduces the required skill level and associated 
production cost. Specific enhancements must be evaluated in terms of the 
Integrated production process on both a single and multimission basis to 
determine the net benefit and reU-ive priority for Implementation. 

An integrated production planning system is one of the keys to 
maintaining a productive workforce and identifying where improvements 
are needed. The availability of a management system tool that proviaes 
the capability to develop accurate work planning schedules, project 
resource requirements, assess program status/evaluate impacts and 
alternatives to changes is essential to the management decision process. 
Continued refinement of the prototype moJel of the JSC flight-to-flight 
reconfiguration process will provide a more in-depth analytical capabil- 
ity to support management in meeting the challenges of the dynamic STS 
operations environment. Long term improvements include the development 
of an expert system application where the management decision process 
based on manual analysis of scheduling- alternatives is automated to 
produce an optimum schedule. 

Authors' Biography 

William A. Johnson is a Project Manager with the Rockwell STS 
Integration Project Office. He received his B.S. degree in Mechanical 
Engineering from Carnegie-Mellon University. His 16 years of 
Apollo/Shuttle experience Includes flight and ground sy-:ems design 
integration, test and mission operations support requirements and 
management systems integration. 

Christian T. Thomasen is a Plans/Schedules Management Adminstrator 
with the Rockwell STS Management Information Systems Office. He received 
his A. A. degree in Business Data Processing/Computer Science from San 
Jacinto Junior College. His 16 years of experience encompasses a 
wide-variety of applications spanning both the commercial and aerospare 
industry including computer operations, data base management, hard- 
wa e/software evaluation and project/performance management systems. 








Robert C. Angler, IBM Corp. 

Leon D. Swartz, IBM Corp. 

Joe H. Wilson, IBM Corp. 


This paper do-uments a fundamental change to the way Space Trans- 
portation System flight preparation is done. It involves. (1) systematic 
restructuring the STS flight preparation task, to minimxze its R&D con- 
tent, (2) development of the STAR System to structure and support the 
remaining work flow. STAR is an integrated software system providing 
automation and quality control mechanisms necessary to create a 
piouuctior.-like process. This approach has been implemented using tools 
and methodology developed by IBM under' NASA contract, to produce payload 
flight data requirements. It offers increased product quality, reduced 
cost per flight, and greater responsiveness to change. 


The Space shuttle was designed for operational flexibility neces- 
sary to take on a variety of missions. Its flexibility is achieved by 
adapting orbiter and ground systems to the specialized requirements of 
each flight. However, the scale of modification is immense. The systems 
which have evolved require extensive software tailoring in order to per- 
form orbiter and ground operations and training. For a typical flight, 
several million individual requirements must be defined, integrated, and 
validated. The flight preparation process to create these requirements 
takes many months and extensive resources. Our ability to produce 
flight-specific requirements in an era of increasing flight rate is a 
major challenge for future STS growth. 

The NASA Spacecraft Software Division was early to recognize end 
respond to this challenge. In 1979, they chartered an IBM task to define 
an approach to flight preparation of Or'oiter Avionics Software in the 
Shuttle Operations era. The goal of this task was to define a way to 
reliably produce mission-reconfigured Flight Software products for a high 
flight rate (over 20 flights per year), at a greatly reduced cost per 
flight [2] . This set of objectives focused attention on potential ap- 
plication of production- line methods to STS flight preparation. 

The task scope covered preparation and verification of the 
flight-tailored Primary Avionics Software System (PASS) load, and inte- 
gration of the orbiter 's Mass Memory Unit (MMU) contents. The former 
includes flight-critical Guidance, Navigation, and Control (GN&C) sys- 
tems, as well as Systems Management and Payload Management systems es- 


'Z"'fi^ -■: 

IrMIi^4|||hA^-.i^' >'«WvK^^^t-'- 


sertial to mission success. These applicacions necessitate rigorous 
control of product quality. 

The Space Transportation Automated Reconfiguration (STAR) System 
was developed by IBM to address this need. It is both a tool and an ap- 
proach, which results in restructuring the R&D tasks of Shuttle prepara- 
tion to form a production process. 


The approach utilizes production methods to re-engineer the tasks 
of flight preparation [1]. Existing methods made extensive use of R&D 
processes and tools, treating every flight as a special case. Flight 
requirements were defined in terms of differences from a previous flight 
baseline, causing simple manifest changes to result in wide-spread 
changes. These techniques could not support the demands of an increasing 
flight rat.. 

A production engineering discipline was applied to the preparation 
of flight-tailored software, in a manner analogous to early automobile 
manufacture. Starting with a hand-crafted prototype (eg. STS-5), 

• re-dfcsign it for manufacturing 

• design and build an assembly line to produce it on 

• establish a supply system for basic parts 

• establish a quality control system for production 

The principal difference is that these tasks were applied to logic and 
data, rather than glass, rubber, and steel. 

Ths product ion -line approach depends on ability to isolate inde- 
pendent components, which presented several technical challenges. It was 
first necessary to separate the effects of flight preparation from those 
resulting from development of new capabilities. Once functional changes 
were separated out, monolithic collections of data requirements remained, 
containing tens of thousands of parameters. A flight preparation process 
had evolved which managed changes to an integrated set of requirements, 
obscuring the origins of its content. 

Secondly, it was necessary to identify component parts. Major 
drivers were identified, which include the launch site, launch date, 
orbiter vehicle, flight trajectory, and cargo characteristics. Iden- 
tification of drivers causing changes in subsets of parameters made it 
possible to divide the data into smaller parts. Each component contains 
a group of functionally related parameters, which change at the same time, 
for the same reason. These exist independently of flight assignment, and 
remain relatively stable in content. 

Third, the rules governing selection and integration of parts had 
to be defined. Grouping of data into components provides leverage for 
flight selection: identification of a higher level name is equivalent to 
many detailed parameter substitutions. Integration and validation func- 
tions can also be automated to simplify flight-related tasks. When all 
components have been selected and integrated, a flight-specific require- 


I ..'I'^C^.V ..^ 





I lautanMMl 

M lX.lHtW 


/ tmJxrntMj/ 




\ WkUFEST / 














inents baseline can be produced. 
Figure 1. 

The full approach is illustrated by 

This approach constitutes a basic change to the way the flight 
preparation task is viewed. Rather than development of a unique fligh* 
requirements baseline, the task becomes collection of f light-iadepende.iL 
components, which are later selected and integrated on a flight basifi. 
This concept allows research and development efforts to focus on provision 
of new components based on projected future needs, while streamlining and 
automating production tasks. Once this capability is in place, available 
components build up until it becomes effective to reuse existing elements. 
A measure of standardization also encourages the reuse of previously 
collected elements. As a result, R&D effort i? minimized. 

Production engineering converts recurring R&D activities to one- 
time tasks, in a structure which encourages their reuse. The task objec- 
tive is changed from development of an integrated product to development 
of components with brotJer applicability, which can then be reliably in- 
tegrated together. This approach considers R&D efforts to be capital 
investments, to be utilized by the STS Program long after the flights for 
which they were performed. 

The Space Transportation Automated Reconfiguration (STAR) System 
is the first major implementation of the production engineering approach. 
It is a pathfinder providing major new capabilities to the STS program. 
It divides flight configuration requirements into separately manage-"- 'e 
components (called "units"), which exist independently of flight »,•. 
It combines quality-enhancing functions of validation, inspection, -id 


^•"SS^ V • 

4», -A ,lU'^*.Vtx ' '^v. 




configuration control into the requirements collection process. It al- 
lows controlled integration of units into larger assemblies; at each step, 
consistency constraints are validated. Flight-specific requirements are 
produced by integrating only the highest levd units needed to define its 
content. This provides the flexibility necessary to respond to the 
changes in flight definition which are inherent to a transportation sys- 


Tlie Space Transportation Automated Reconfiguration (STAR) system 
is a software configuration and data management system developed by the 
IBM Corp. under contract to NASA. The initial delivery in June, 1985, 
provided the basic capabilities needed to colleC;, validate, and inte- 
grate Space Shuttle Systems Management and payload data into the overall 
reconfiguration production process. This data includes the definition 
of payload related telemetry streams, uplink commands, onboard displays, 
and the onboard processing required to command and control the payloads 
themselves. The first flight using products from the STAR system is 
STS-61E which is scheduled to fly in early 1986. Subsequent planned up- 
grades to STAR will add additional payload processing capabilities plus 
incorporate data management support of the vehicle parameters monitored 
by the on-board General Purpose Computers (GPC's). 

The development of STAR began in late 1982. IBM was responsible 
for all aspects of the software engineering task: requirements analysis 
c'nd specification, implementation and integration, and coordination of 
the customer's user acceptance testing and release to operational pro- 
duction. Over 149,000 source lines of PLl and ADF code were delivered. 
At the time of release, NASA engineers had already input over 22,000 
payload parameters and 55,000 description definitions defining 54 pay- 
loads. The transition to production was smooth and the system readily 
accepted because the user community was very involved right from the be- 
ginning. Therefore useability received as much consideration as 

Purpose of STA R 

Several years ago NASA realized that the reconfiguration data 
management tools and techniques r.hey were using would not support their 
future Shuttle flight frequency goals. The process was labor intensive 
and error prone because of of the amount of data involved end lead time 
required. Additionally, there v;as no convenient means of reusii^.g previ- 
ously defined data, so each Kiission required a "custom" system put to- 
gether by a team of "experts". The impact of errors was frequently 
significant because they typically weren't caught until late in the 
critical path processing for a flight. STAR is one of several new tools 
intended tc eliminate Lliose problems and deficiencies. 

The purpose of the STAR system is to automate, centralize, and 
control the collection of Shuttle reconf igurable data that is processed 
by the ground find on-board flight computers. The remainder of this paper 
describes the concepts and approach chosen by IBM, and the features and 

'f capabilities of the STAR system itself. 









MIS > 


C* ■ 

Cir9« Biy 

CE a 

Orgs Elamnt 



n. ■ 


VEH a 


ve ■ 

Vchicl* ElMwnt 

VS$ > V*l<kt« Subly'itx 

Basic Concepts 

The basic STAR concepts are: 

• Software enforced configuration control 

• Software enforced data quality/error detection rules 

• Automated selection for reuse 

• Streamline and control mission manifest tasks 


The STAR system enforces these basic concepts via: Data Structure, 
Data Integration, Data Control, User Access Control and Data Partition- 
ing. These terms are introduced below. Additional detail on each topic 
is provided in subsequent sections. 

Data Structure is controlled by use of a predefined hierarchy, 
shown in figure 2. This hierarchy segregates each group of reconfigura- 
tion elements into controlled structures called "units". The units used 
in STAR include: Mission, Cargo Bay, Cargo Element, Payload, and Display. 
Data is controlled by only allowing data in a "working" state to be 
changed. Data may only be used for a reconfiguration after all audits have 
been passed and the data has been "baselined" for use. Access to the STAR 
data bases are limited to a set of predefined users. The tasks that a 
user can perform are further restricted by user function (e.g. data entry 
clerk, data suppliers, data coordinators, mission managers, approval 
board chairperson). Data integration is enforced at data entry for audits 
that detectable on a single screen of data (called a list or data cate- 



gory). Audits that cross screens of data are verified by manually initi- 
ated batch processes called integrators. Each unit type has an integration 
processor. All screen audits and integration checks must be successfully 
passed before data can be submitted for baselining. Data collection is 
performed mission independently to allow data reuse on multiple flights 
(e.g. parameters required by the Pay load Assist Module (PAM) upper stage 
are stored once and reused on each flight). Mission dependent data values 
(mission manifest, payload bay address, etc.) are collected in a mission 
independent data base which contains all known missions. A mission de- 
pendent data base is created for each mission to allow the mission inde- 
pendent units to be configured with the mission dependant values. 


Data Structure 

The data ^tructnre of i"hp. STAR system is hierarchical and is con- 
trolled via a template called a prototype. Collection within the proto- 
type is further controlled by data categori«s, list categories and units. 
These categories define ♦ihe content, conbLiaints and structure. 

DA'iA CATEGORIES con .rain actual data values which define or share 
i some common purpose or function. For example, the calibration coeffi- 

cients for a parameter would be collected in a single data category. Each 
.| data category has a corresponding specifically designed input data 

' screen . Each data category a 'so has a set of intra-data category audits 

-J which are enforced at data entry. These audits must be passed before the 

"? data can be baselined. Examples of audits include value, format, range and 

2 inter-value constraints. 

LIST CATEGORIES contain REFERENCES to other list and data catego- 
' ries. List categories are used to group multiple parameters with a common 

source or function together. Examples of a list category include the 
references to the set of payload measurements and payload commands for a 
payload. The list categories are used to construct a template called a 
PROTOTYPE . The prototype defines the lists and data categories that 
may be referenced within the specified prototype. In addition, the pro- 
totype defines the number of list categories that may be referenced in 
the specific prototype. For example, the Cargo Bay prototype defines the 
maximum number of Cargo Elements that may be defined in the Cargo Bay. 
Any number of specified data categories may be referenced within the 
prototype. For example, many parameters may require calibration in a 
payload. The filled in structure of a prototype is called a UNIT. The 
Space Shuttle payload units are: Display (an onboard display de- 
scription), Payload (an upper stage, pallet or spacecraft on an upper 
stage or a subset of a large payload like Spacelab), Cargo Element (a 
collection of Payload and Display units which are manifested as a group), 
Cargo Bay (the Cargo Elements and Displays (which c.nntain multiple Cargo 
Element data) which make up a Shuttle Cargo Bay), and Mission (the Cargo 
Bay and Space Shuttle Vehicle which will fly on the specified mission). 


■•} References are made via a standard name structure consisting of 

t the category name, occurrence name and a qualifier. The category name 

^ selects the specific category definition. The occurrence name defines a 

• specific instance of the category. The occurrence name may have specific 

I audits. The qualifier is used to distinguish between concurrent versions 

of the occurrence name. Qualifiers allow different models of a parameter 



.i^.iL-XixXl.^.^'lS. .''^j 

or unit to be maintained concurrently in the data base. For example, an 
upper stage may have a new model which has unique parameters from the old 
model. The new model may reference all of the unqualified common parame- 
ters and the new unique qualified parameters. The old unit would continue 
to reference the unqualified parameters. 

Data Integration 

Some audits cannot be performed online because they exceed allow- 
able online computer resources, validate inter-^category audits, or pre- 
? vent units that contain invalid inter -category data from being submitted 

for review. Thei:e audits are grouped into batch processes called 
integrators. Each unit has a dedicated integrator. Special data catego- 
ries exist to allow batch job submittal, the review of resulting 
error/warning messages, and to view other relevant data collected by the 

* • Data Control 

': i All changes to data are inventoried under an authorizing control 

^j-i in.^trument called a Data Change Request (DCR) . An authorized user must 

fi create (OPEN) a DCR before any data can be changed. Ths unit or units to 

be changed must be selected or a new unit may be defined using the pro- 
totype. Within each unit the author must authorize the set of categories 
to be changed. Other authorized suppliers of a category may change or 
define occurrences of the authorized categories. All changes ere con- 
tained in change records called WORKING COPIES. All working copies must 
pass all defined audits before the occurrence can be submitted for review 
(FROZEN). Before a DCR can be submitted for review all occurrences must 
be frozen and vhe authorized and supplied categories must match. The 
control board must then approve the DCR and the batch process which 
baselines the data must be executed to change the working copies from the 
frozen to baseline ^-tate. 

Access Coiitrol 

Access control ic enforced by user identification, user function 
and other restrictions. Orly authorized users may log on. Each user has 
a unique ID which is used to tag each change made by that user. Users are 
organized into functional groups. These groups are: configuration control 
(maintains access control data), data suppliers (change data values), DCR 
coordinators (control DCR submission), board chr^irperson (dispositions 
DCRs), reviewers (browse data), product generators (perform production 
product generation), and mission .-nanagers (only authorized supplier in 
mission dependent data bases). A separate access control data base further 
restricts users within the above general groupings. This data base defines 
users, alternate users for each user, categories, data suppliers for each 
category, DCR coordinators, qualifiers and mission managers. 

;? Data Partitioning 


I Data collected in STAR is partitioned into mission dependent val- 

I ues and mission independent values. Mission dependent values are col- 

, lected at a high level. This allows mission independent data to be reused 

* on multiple missions. In STAR, the Display, Poyload and Cargo Element 


•k iM(Mi^j^:^'\^.:' 

Units are collected in a mission independent manner. This allows standard 
upper stages to be used on multiple missions without reentering the data 
values. Data is separated into units with this goal as a major consider- 
ation. For example, the PAM uppei stage is broken down into several Pay- 
load units differentiated by whether or not the parameter is for standard 
upper stage, optional upper stage services, upper stage spacecraft ser- 
vices or upper stage Orbiter software generated parameters. 

This separation of data allows NASA to easily update mission man- 
ifests when changes are required. Only the Cargo Bay unit needs to be 
changed to define the new mission dependent data value set. On the Shuttle 
the mission dependent data values include the data buss addresses, data 
buss device type, display numbers, flight software load, IDs which dif- 
ferentiate between multijple payloads of the same type (e.g. multiple 
PAMs) and other similar values. 

Since this data is maintained in a mission independent data base 
users viewing the mission independent data do not "see" the mission de- 
pendent values on the same screen as the mission independent values. A 
mission dependent data base is created for each mission to capture a 
complete set of data for a specific mission. The mission dependent values 
are "buried" in the low level occurrences so that users viewing the data 
can see mission dependent and independent values simultaneously. Changes 
are restricted to mission managers in the mission data base. Only make 
work mission unique changes are intended to be entered directly into the 
mission data bases. 

The payload data is provided to the various facilities that use 
payload data via a standard transfer data set called 'x Payload Data 
Transfer Format. These facilities include the flight software, mission 
simulators and the launch facility. In some cases the additional 
channelization data is added to the data produced by STAR *■; another el- 
ement of the reconfiguration system prior to delivery to the user. 


The STAR system implements a production engineering approach which 
minimizes the effort and errors encountered in building mission manifests 
for the Space Snuttle payloads. STAR accomplishes this by providing: 

A flight -independent catalog of data components 

Enforced configuration control of all data changes 

Enforced audits which validate data on entry 

Data reuse by selection and integration of catalogued components 

Tools to aid the manifest/remanifest process 

STAR also controls reconf igurable Space Shuttle vehicle parameters that 
are required to be monitored by the onboard general purpose computers. 
STAR is currently being upgraded to also process all reconf igurable data 
vfllues for the Orbiter Guidance, Navigation and Control computers. When 
this is accomplished in the spring of 1986 STAR will contain generic 
functions which will allow control of any data. The approach taken by 
STAR is general, and can be applied to all areas of STS flight prepara- 





[1] Angler, Robert C, "Organizing Space Shuttle Paranetric Data for 
Maintainability", Journal of Guidance. Control, and Dynamics . Vol. 
6, No. S (September, 1983), pp. 407-413. 

[2] NASA, JSC- 16763 Software Production Facility Operations Planning 
Document, Vol. 1, Book 1 (December 1981). 

Robert Angier is an Advisory Systems Engineer in Space Systems 
Software Technology, IBM Federal Systems Division, Houston, Texas. He 
lead concept formulation and initial definition of the STAR System. 

Leon Swartz is Manager, IMS Data Base Applications Development, 
IBM Federal Systems Division, Houston, Texas. He managed the development 
effort which produced the STAR System. 

Joe Wilson is an Advisory Systems Engineer in the Reconfiguration 
Systems Development organization, IBM Federal Systems Division. Houston, 
Texas. He was technical manager of the Initial release of the STAR ap- 









^^8 6-15163 


George Mamderlink 

Lawrence P. Clark 

Willj,»i M. Bernstein 

W. Warner Burke 

Teachers College, Coliiit)ia University 


Ihis study examined the relationships antxig manager behavior, 
group climate and nonagerjal off ectiveness . Survey data were collected 
fztxn 435 Q414-15 managers and their subordinates at NASA adeeming 
management practices and perceptions of the group environment. Per- 
formance ratings of managers were obtained from their superiors. Tt-ie 
results strongly supported a causal model in v^ich subordinates' 
climate perceptions mediate the effects of manager behavior on perfor- 
mance. Ihat is, the development of group climate provides the prTcess 
through v*iicii the effects of manager practices may be understooc 
Analyses also revealed that the function performed by a manager and 
his group (e.g... research) influenced the specific nature of the 
causal dynamics. Sane inplications of the results for management 
j training and developnent are discussed. 


1 Social science research has shewn that the behavior of leaders 

'■ and rtanagers is a potent detpjminant of group motivation and behavior 

[4;5] . Moreover, receiit research has shown that both a manager's cwn "^^ -^ 

1 sense of his behavior, and the extent to which his subordinates share ^ 

j his perceptions of his behavior must Le considered in order to explain 

ii a manager's effect on his grtjup [1] . j 

I A manager's view of himself, and the congruence between his 

view and his subordinates' view of him, operate to effect group behav- 
i ior by influencing various perceptions and expectations held l^ subor- 

I dinates about the group envirc»inent , C^iite sensibly the variables that 

j have been found to be useful for prediciting the achievement and social 

' behavior of groi^js include: 

(1.) the extent to vAiich subordinates perceive their goals, 
tasks, ctnd roles cleeurly 


•JLii^M^X^><H\''^*'-'^-~' - - \^ 

(2.) the extent to which subordinates believe that nigh, 
challenging standards are beijig used to evaluate their performance 

(3.) the extent to which subordinates expect to receive 
evaluative feedback about their performance and to have ii^t into 
group decision n^ing (i.e., the extent to w'.idi suboicUnates expect 
to participate in group evaluation and decision making) 

(4.) the extent to which subordinates expect to be r^iarded 
mcv)etarily and with praise for achieving performance^ standards 

(5.) the extent to which suborditates expect that affiliating 
with other group members will result in their being rewarded with 
positive social feelings of friendliness, supfx rt, and trust 

(6.) the extent to v*iich subordinates believe their gnxp 
has c^jen, cooperative relaticais vith other orgauuzational tinits 

Taken together, tJiese perceptions of the group enviroranent which re 
important predictors of task and social b^iavior have been callt- 
group climate. 

In sum, social science research eind theory indicate th-it a 
group's performance may be explained euxl predicted tliroia<^ the use 
of a general causal model (see Figure 1) . This model assigns group 
climate perc^tions a central, mediating role. Self- and fjubordi- 
nate perceptions of manager behavior influence climate whicJ\ then 
determines the quality of the group's task and social outoan<?s. The 
formation of group climate perceptions thus provides the process 
through which the irotivationa] aix?. effects of management 
practices can be understood. 

Figure 1 
Schematic Model of the Climate Mediation Process 










The present study represents an application of the 

practice climate idea to the realm of manacer devsloginent at 

NASA. In particular, the findings to be discussed indicate the way 
in which manager behavior and subordinate cli'nate perceptions effect 



^ s 

* I 



' i 


the perc^tions a manager's sv^ierior has about the rtenager's oon- 
petence. In some sense, of course, a superior's perc^>tions of a 
group's manager may be related to group performance. But the models 
tested here do not include groqp performance measures. The prime 
usefuliiess of these mcxiels resides in their ability to map the 
dynamics controlling the ocnnunication of performance relevant expec- 
tations and perceptions between mEUiagers, their subordinates < and 
their si;5)ervisors. 

A furtlier aim of the pre3ent study vias to determine vAiether 
group climates most advantageous for manager perfomanoe ratings 
differ by functicwi within N?^SA. That is, different aspects of the 
group envixcaTnient may facilitate better performance cutoomes and 
social relatirais depending upon the nature of the group members' 
functicffial respcMisibilities . In terms of the prc^xjstjd model, this 
dependency woulc' be manifested in different climate fjerceptions 
mediating the effects of manager behavior on performance appraisals 
in different functions. 



The first step in assessing the causal dynamics prevailing 
vvithin functions among managenent practices, group climate percep- 
tions, and superiors" performance evaluations was to measure each 
of these variables, "i^is task was begun by collecting data about 
perceptions of nanager behavior and group climate fron 435 middle- 
level (CM.4-15) managers and their subordinates at NASA. More 
spec^''ica''ly, managers and subordinates were asked to rate the 
extern, to vtiich 80 different nanagement practices and 42 different 
climate expectations prevailed in their groups. 

Factor Analysis of Practices and Climate Perceptions 

Given the large number of management practices and climate 
perceptions assessed, factor analyses were performed to reduce the 
ocnplexity of data within these danains. Accordingly, two factor 
analyses were conducted, one for manager perceptions of their own 
behavior, and another for group perceptions of climate. Ihe goal 
of these analyses was to represer.t the rather large variable sets 
in terms of smaller groups of hypothetical variables [3] . These 
underlying factors can be interpreted as organizationally shared 
schemes for organizing thoughts and perceptions about the NASA 

A principal ccqponents factor analysis with oblique rotation 
was ajplied to managers' ratings of the extent to which they perform- 
ed each of 8C different behaviors. A set of 10 factors was found to 
reflect the dimensions along \*iich NASA managers perceive their cwn 
behavior. These were: PraitDtang Achievement, Monitorina Projects, 




Identifying with the Organization, Teiking Others Perspective, Deal- 
ing with Performance, Creating Tciist, Involving Others,, Recognizing 
Others, Managing Resources, and Dealing with Problems. A self- 
rating scale was then derived fron each of the factors. High scores 
on these scales (range: 1-5) mean that a manager sees himself as 
performing more of the definitional behaviors. 

In order to assess agreement between a manager and his 
subordinates oonceming the manager's behaviors, subordinates' 
ratings were averaged over each of the items cotrprising a factor, 
and then subtracted fron the itanager's self-rating. Positive 
difference scores then indicate that the nenager has rated his 
behavior more favorably than his subordinates have rated it. 
Negative difference scores indicate that the subordinates have a 
more favorable view of the manager's behavior than the manager 
has of his own behavior. 

A principal ccnponents factor analysis with oblique 
rotation was also applied to subordinates' perceptaons of their 
group climate. The results revealed that* group climate perceptions 
were found to vary along six basic dimensic»is: Clarity (of goals 
and tasks) , Getting the job done. Participation (in decision- 
making and performance evaluation processes) , Standards (level of) , 
Social Rewards, Interunit Relations. Six group clinvte scales 
vvere created based on this solution. 

Performance Evaluations 

Performance ratings of managers were then obtained fron 
their superiors. The system used to derive these performance 
ratings is outlined in the NASA Supervisory and Managerial Perfor- 
mance Rating System (1980) . This systan uses a behaviorally 
anchored rating technique; detailed guidelines specif^' hrw perfor- 
mance objectives should be develc^aed, evaluated, and reviewed. In 
order to use performance appraisal ratings as an outoone variable 
in the planned analyses, managers were assigned points based on 
their ratings each year over a three year period (FY 1980-1982) . 
Ihe following method of scoring was used: Managers were given 10 
points for each outstanding rating received during the FY 1980-1982. 
They received five points for every highly successful rating re- 
ceived, r-nd one point for every successful rating. Managers were 
not given any points for satisfactory ratings. No managers in this 
sample received unsatisfactory ratings. 

Both the set of 10 nanager self -perception variables and 
10 itanager-subordinate perceptual difference variables were includ- 
ed as predictors of climate and performance appraisal outcones in 
the subsequent evaluation of the basic causal model. Scores on 
J the group climate scales were tested as variables. These 

I analyses we e conducted seperately for work groups having different 

* functional responsibilities. Each group was classified according to 

the manager's reported function. This resulted in a four-way 
■ classification of groups: engineering, research, project managarsent 

V I 






and adninistrative/anesource. 
Path Analysis 

In ozxSer to dencxistrate that groap clijnate perceptions 
mediate between pexxxptians of manager behavior and evaluations 
of his performance three requirements must be satisfied [2] : 

(1.) a nanager's perceptions of his own behavior and the 
extent to v^ch his subordinates agree with those perceptions affect 
his performance rating. If these effects do not exist, thei« is 
nothing for group climate peroepticais to mediate. 

(2.) manager and subordinate peroepticxis of manager 
practices influence group climate. Cliiratc perceptions can only 
mediate the effects of manager behavior variables if they eure 
themselves affected. Individual climate variables vrfiich pass this 
test can be considered "potential" mediators of manager practice 

(3.) group climate porc^tions must influence superior's 
ratings of iranager performance when the effects of manager and 
suborxlinate perc^Tticais of manager behavior are controlled. Satis- 
faction of this requirement r^resents the actual mediating effect 
of groap climate. 

If these three requirements are met, mediation can be 
claimed. That is, the eff3c:s of manager b^avior variables cwi 
performance evaluations can .-je understood in terms of their in- 
fluence on subordj-iates ' perceptions of group climate. 

To test these requir^jnents for mediation, path analytic 
techniques were used. Path analysis is a type of imltivariate 
statistical method for testl-vg causal inferences with correlational 
(or survey type) data [6] . First, superiors' ratings of manager 
perfomance were regressed cxi the 10 manager self -perception 
variables, and the corresponling 10 manager-subordinate perceptual 
difference variables. Next, each of the six climate variables was 
regressed on these 20 manager bdtiavior variables. Finally, 
si^jeriors' ratings of nanager performance were regressed on the six 
liitBte variables and the 20 nanager bdiavior variables (i.e., the 
full path model) . Entering all six climate variables into this 
equation makes it possible t-i evaluate their relative mediating 
effects. Ihese three equati-ins were estimated for each of the four 
fuxictions in order to examine: whether the causal dynamics among 
practices, climate, and perf-irmance-related outcomes differ in 
relation to the primary task performed by the manager's group. 

A step-wise regressi-jn procedure was used to test the 
effects of the multiple predictors on each of the climate var- 
iables and si:?)eriors' perfomance ratings. Variables were entered 
into the regression equation if the p-value associated with the 
variable's path coefficient fbeta weight) was less than .10. It 


^ '>:- A 


is these 5inalyses which establish the causal chain in v^iich per- 
ceptions of managerial practices affect group climate perceptions, 
which in turn affect superiors' ratings of a manager's performance. 


An examination of the path models for each of the four 
f'jnctions indicated that groc^ perceptions of climate did mediate 
the effects of manager behavior perceptions on superiors' ratings 
of manager performance. In addition, a nanager's perception of 
his own behavior and the extent to v*iich his subordinates agreed 
with those perceptions affected performance evaluations directly. 
Ihat IS, the effects of seme manager behavior variables were not 
mediated by group climate perceptions. The results also suggest- 
ed tliat not all group climate perceptions eire relevant in deter- 
mining a superior's evaluation of the grcAiy manager's performance. 
Precisely- which group climate perceptions play a mediaiting role 
depends on the function performed by the manager's group. 

llie path models for the research and engineering groups 
are presented in Appendix A to illustrate some of the major as- 
pects of tiie present analysis. Only statistically significant 
relations between predictors and the nediating and outcore 
variables are represented in the path diagrams. Both the research 
and eigineering path models predicted meaningful variation in 
perfoiirance ratings that cannot be attributed to chance (Research, 
18%; Engineering, 11%), indicating that the models have predictive 
validity . 

Research Groups 

The full path model for research groups is displayed in 
i^pendix A, Figure 1. These results demonstrate that superiors' 
evaluations of manager performance are directly af l -ed by two 
factors, the perceived climate for participation ai. subordinates 
and a manager's self -perception of the extent to vAiic . he involves 
his subordinates in group planning and decision-making (Involving 
Others). Specifically, a manager's performance rating is higher: 

(1.) the more his subordinates expect to participate in the 
areas oi decision-making and perfontance evaluation (path coef- 
fiijient=.32) . 

(2.) the more he perceives himself as creating opportunities 
for his subordinates to become involved in the planning of work, and 
to have influence during the decision-making process (path coef- 
ficient=.27) . 

Taken together, these results suggest that recognition of 
effective management or research groups at NASA is based on a man- 
ager's ability to catmunicate a participative approach to his sub- 




ordinabes and supervisors. Manager's v*io are able to dcvelc^ a climate 
high in participation are recognized by their superiors as Tore out- 
standing performers. At least tMO hypotheses can be suggested to 
explain this relation: 1.) research groups characterized by higher 
levels of participation actually achieve more and have better social 
relations, and 2.) those v^ evaluate the performance of managers of 
such grot^ hold the "iirplicit theory" that participation by subordi- 
nates is iiroortant in attaining these positive cutocnes. Of course, 
these hypotheses are not mutually exclusive; eis discussed previously, 
si^jerior's notions of vAiat leads to good performance cu:e likely to be 
' correlated with the actual determinants. 

The direct effect of Involving Others on performance evalua- 
ticais provides further evidence that participative behavior on the 
part of the manager is particuleurly valued by his superiors. More- 
over, by allowing group mentjers to contribute to planning and deci- 
sion-making processes, a manager may actually be ocnnunicating to 
! those outside his group that his subordinates cure extremely carpetent. 

{ As noted, the path model indicates that the evaluation a 

j nanager receives fran his superiors can be increased by raising his 

j subordinates' expectaticsis about participaticai. In fact, 53% of the 

' variation in subordinate participation expectations can be attributed 

; to manager behavior perceptions. Hence the four paths to the climate 

; variable suggest hew a manager may proceed to raise th.ese expectations. 

' Participation is expected to be hi^er the more a itanager sees himself 

as creating tnist (i.e., building supportive relationships with sub- 
ordinates, emphasizing coc^jeration; .30), and the more willing he is 
to involve subordinates in planning/decisicm-making (.33). With 
respect to the Involving Others variable, the perceptual agreement 
between a manager and his subordinate is a stronger determinant 
(-.66) of Pcirticipation than the manager's self -perception. It 
appears that subordinates" participation expectations are higher the 
less the manager's estimate of his willingness to involve others in 
planning exceeds his subordinates' ratings of his behavior in this 
regard. In other words, a climate of participation is undermined 
v*»en group manbers perceive their manager making inauthentic claims 
about his attempts to solicit their opinions and include them in the 
unit's planning process. 

Manager-subordinate perception differences on the Recognizing 
Other a ilimension also determine group perceptions of participation. 
Specifically, participation expectations will be higher tl-iG less a 
manager's claims for his recognition behaviors (i.e., providing in- 
formal feedback to subordinates, taking a personal interest in sub- 
ordinates) exceed those of his subordinates (-.26). 

Engineering Groups 

Ihe analysis for engineering groups indicates that subordi- 
nates' climate perceptions concerning both participation and inter- 
^ unit relations mediated managenent practice effects on the outccme 




tr^^ES^--.-- , 



varie±)le (see AKsendix A, Figure 2) . Self- and subordinate percep- 
tions on two management prac±ice variables also directly affected 
superiors' evaluaticais. Specifically, Figure 2 shows that a manager's 
performance rating is higher: 

(1.) the more his subordinates expect to participate in the 
areas of decision-making and performance evaluation. (.23) 

(2.) the less his subordinates perceive their relations with 
other organizational units as cpen and ooc^jerative. (-.23) 

(3.) the less he perceives himself as involving others in 
planning/decision-inaking processes. (-.18) 

(4.) the largfj: the difference between the manager's evalua- 
tion of his participative approach and his subordinates' evaluation 
of him. (.32) 

(5.) the more he perceives himself as identifying with 
organizational goals and dajectives. (.28) 

(6.) the less his pierception of tlie extent to viiidi he 
identifies with the organization exceeds his subordinates' percep- 
tions of him on this dimension. (-.35) 

As was true of those supervising research groups, engineer- 
ing managers v*io are able to effect a climate high in participation 
are evaluated more favorably by their superiors. These evaluations 
may be based on relatively objective performance criteria with groups 
characterized by an atmosphere of participation actually performing 

"The extent to v*iich subordinates expect to have input in 
decisionTTiaking depends heavily upon several aspects of their man- 
ager 's behavior. Managers v*io see themselves as understanding of 
others' point of view (Taking Others Perspective) will increase sub- 
ordinates' for participation (.18). However, the 
manager's effectiveness in this regard is constrained by the extent 
to v*iich his subordinates share his perceptions about the meaning of 
his behavior. If the manager-subordinate difference is large, 
particpation is lowered, as evidenced by the negative path coef- 
ficient (-.31). Several other aspects of the manager's behavior 
also influenced performance expectations. Ihese variables were also 
found to affect the climate for participation in research groups and 
do not require further discussion here. 

Although group climate perceptions of high participation are 
likely to enhance a manager's performance rating, tlie manager's cwn 
participati.on behaviors seem unappreciated by his superiors. That is, 
the more a manager perceives himself as involving others in decision- 
making, lower is his performance evaluation (-.18). Apparently, the 
superiors of engineering group nanagers feel that a directive approach 
is more appropriate than a more participative style. These results 


^-^ESS' •-.-' . 



suggest that a mana<jer's performance evaluation is likely to be better 
to the extent that he is seen by his superiors as usijig a ncsipartici- 
pative approach with his unit. JUst the opfxjsite iitpressicn, however, 
should be mar::iged with ones' subordinates in order to enhance perfor- 
mance ratixigs. 

Interunit functicxoing, the seoraid madiator of nanagement prac- 
tice effects, iias a negative iirpact on superiors' evaluations of man- 
ager perfomance (-.23) . Interestingly, this is the only instance in 
which the "'more is better" rule does not e^ly with respect to a cli- 
nate mediator, i.e., more positive pero^Jtions of a unit's relaticais 
with other groups are related to lower performance ratings. It appears 
that "good interunit" relations or "oocperative relations" are not 
valued by the superiors of engineering group itanagers. The prevailing 
notion may be that corpetition among cfrxDups and accsupanying feelings 
of territoriality and distrust are more potent motivating forces than 
ooc^jerative tendencies. Managers v*x) are perceived to be able to 
maintain a basic level of insecurity are benefited given this value 
orientation (i.e., ccnpetiticxi over cooperation). 

The model indicates that a manager's ability to create trust 
and respect among his subordinates, and his recognition bdiaviors are 
infjortant determinants of ijiterunit functioning. To illustrate, if 
a manager perceives himself as interested in his subordinates and as 
providing recognition, and if this interest is seen as genuine by the 
subordinates themselves, better inter-group relations result. These 
results suggest that good "inter-group" relations are facilitated by 
good "intra-group" relations. 

The final set of effects to be discussed concern an erigineer- 
ing manager's perceived identificaticai with the organizaticxi. The 
meaning of these results seem apparent. To the extent that managers 
identify with the organization, enphasize aooonplishing the work of the 
organization and have good relations with upper level executives, they 
will be awarded high performance ratings. In short, "team players" 
are valued among those in engineering. 


Itie causal models produced by the present study (and other 
similar models) have been used as part of the NASA Management 
Education Program (MEP) conducted at the Wallops Island Training Cen- 
ter. An inportant part of the two week MEP training program involves 
feedback to participant uanagers about their subordinates' perceptions 
of the itanagers' behavior, group climate, and the differences between 
manager and subordinate perceptions. 

In March 1985, feedback to managers was acconpanied by presen- 
tations of the models generated by the present study. Since the usual 

ri 59 


feedback process provides managers with information about a large set 
of different behavior ratings and climate percepticxis, tho. models can 
vork to focus pcirticipcints' attention cm the practice and climate di- 
mensiox^s that may be most critical for explaining the memagers' 
effects on others. 

Ihe 30 March MEP participants were divided into four groups 
according to functional area. One MEP trainer was assigned to each 
group (research, engineex'ing, project, and administrative) . The 
trainers first described the nature of the results to the group. 
Then, the managers were asked to oonmerit upon the face validity of the 
results, i.e., "Did the model seem to be a sensible representati(Xi of 
influence procfcsses within their cwn group?". By and large, managers 
reported that the mcdels were not inconsistent with the way they per- 
ceived practice to climate effects. And, the models tended to make 
some effects of their bebiavior more salient or explicit. 

Managers were then asked to try to relate their own personal 
feedback results (given in terms of NASA 0114-15 norms) to the models. 
For exanple, the administrator model suggests that subordinates' ex- 
pectations of receiving social reward? are increased to the extent 
that managers are seen as performing three types of behaviors (per- 
spective taking, recognizing others, and involving others) . Adminis- 
ttrators interested in increasing subordinates' social reward expec- 
tancies were advised to check their ratings on these practice dimen- 
sions to locate weak areas, or areas that might be iitproved. By 
focusing the attention of certain iranagers on specific practice areas 
that likely affect ti"ie perceptions of others, training beccmes more 
focused and, hence, practical. 

The notion that "ore's behavior effects others" can be trans- 
formed through this type of work. The change is from a general, 
rather banal, "law of social behavior", to many more precise, instru- 
mental percepts to guide and energize change. 

Work with both model generation and training afplication tech- 
niques is now proceeding. We feel we have only begun the process of 
increasing the iitpact of rigorous enpirical social research on manage- 
ment and organization developnent. 

George Manderlink (PhD, Columbia) is an adjunct professor and post- 
doctoral fellow in organizational psychology at Teachers College, 
Columbia University. 

Larry Clark (PhD, Syracuse) and William Bernstein (PhD, University 
of Texas, Austin) are both organize cion consultants. 

W. Warner Burke (PhD, University of Texas, Austin) is Professor of 
Psychology and Education at Teachers College, Columbia University. 
He has consulted with a variety of organizations, including NASA, and 
is the author of the book Organization Developnent ; P rinciples and 
Practices . 




' z) 

Hmiaqer behavior preoption* i 

Group cll« it« p«rc»ption» i Sup«rior«' prceptioni i 


M'8 Rating 


M-S Difference 

H'a Rating 



M-S Difference 

Manager behavior perceptions ; 


M's Rating 

M-S Difference 

M's Rating 

M-S Difference 

M's Rating 
M-S Difference 


M'S Rating 

H-S Difference 

M's Rating 

M-S Difference 

Group climate perceptions ; 

. t\ 



Superiors' perceptions : 



M's Rating - Manager's perception of himself 

M-S Difference - Manager-Subordinate perceptual difference 




[1] Bernstein, W. M., and Burke/ W. W. Detenninants of group 
cliinate perceptions. Colunbia University, 1984. 

[2] Judd, CM., & Kenny, D. A. Estimating the Effects of Social 
Interventions . Cambridge University Press: Cambridge, MA, 

[3[ Kim, J., & Mueller, C. M. Introduction to Factor Analysis . 
Sage Publications: Beverly Hills, CA, 1983. 

[4] jjewin, K., Lippett, R. , & White, R. Patterns of aggressive 
hehavior in exDeriroentally creexted social climates. Journa l 
o f Psychology ,' 1939, 10, 272-299. 

[5] L-ltwin, G. H., & Stringer, R. A. Motivaticai and Organizational 
CJ imate . Harvcird Business School: Cainbridge, MA, 1968. 

[6] Wright, S. Path coefficients and path regressions: Alternative 
or ocuplementary concepts. Dionetfrics, 1960, 16, 189-202. 




. Lee Avant. Federal Express Corporation 

Robert W. Boozer, Memphis State University 


This paper investigates the Impact of a formalized mentoring 
program as a communication channel for enhancing Information 
distribution. Innovation, and productivity. Formal and Informal 
approaches to mentoring are discussed. Interviews with 11 members of 
\ formal mentor-protege teams Indicate communications In the mentoring 

relationship can affect Individual and organizational Innovation and 


A major premise of this paper Is that the U.S. aerospace program 
Interacts with an environment that has been described as post-Industrial 
society - an environment characterized by Increasing knowledge, 
complexity, and turbulence. Such an environment places special demands 
on organizations; more rapid, frequent, and complex decision making, more 
rapid and frequent Innovation, and more continuous, wide-ranging, and 
directed Information acquisition and distribution. To be effective In 
such an environment requires that organization structure and process be 
designed to reflect these demands. [3] 

The purpose of this paper Is to Investigate the utility of one 
sUernative for meeting some of these demands. In partU 1ar, this paper 
investigates the utility of a formalized mentoring program as a 
coMnunlcatlon channel for enhancing information distribution, innovation, 
and productivity within an or.inlzation. 

Mentoring - The Informal Approach 

As traditionally conceptualized, mentoring is an informal 
relationship between two people - one a senior and more experienced 
individual (mentor) and the other a more junior and l£:s experienced 
-{ Individual (protege). The term for the relationship - mentoring - 

H derives from Greek mythology. As the story goes. Odysseus entrusted the 

!| education of his son Telemachus to a trusted friend. This friend Mentor, 

I became responsible for tutoring . sponsoring, and coaching his protege, 

Telemachus, while Odysseus was away from home. Similar relationships 



\\ ' 


have continued through history as seen In doctor - Intern, master - 
apprentice, and teacher - student relationships. 

Recent research Indicates some consensus about the characteristics 
of mentoring relationships. First, they Involve a number of roles 
similar to those performed by Mentor. K:am [6], for example. Identifies 
five career roles and four psychosocial roles. The five career roles are 
those that enhance career development. These Include: (1) nominating 
and supporting the protege for promotion and advancement (sponsorship); 
(2) assigning responsibilities which bring the protege Into contact with 
key organization figures (exposure and visibility); (3) assigning work 
that helps develop the protege's technical and managerial skills 
(challenging assignments): (4) providing guidelines and feedback about 
work beliavlor (coaching); and, (5) shielding the protege from criticism, 
adverse publicity, etc. (protection). The four psychosocial roles are 
those that enhance the individual's sense of self-esteem, identify, etc. 
These roles Include: (1) providing a set of values, beliefs, and 
behaviors for the protege to follow (role-model); (2) providing mutual 
liking, mutual respect, and positive feedback about performance 
(acceptance and confirmation); (3) creating a supportive climate where 
the protege can discuss anxieties, fears, and conflicts that Interfere 
with productive work behavior (counseling); and (4) establishing a mutual 
relationship of liking, understanding, and Informal social exchange 
(friendship) . 

A second characteristic Is the dynamic nature of mentoring 
relationships; they evolve through a number of phases similar to other 
human relationships. For example, Kram [6] Identifies four phases: 
initiation, cultivation, separation, and redefinition. 

A third characteristic is the informal nature of mentoring 
relationships. That is, the relationships develop without specification 
and guidance from the formal organization. Rather, the relationships are 
influenced by factors found to influence other types of informal group 
and social network formation [13]. For example, the initial phase is 
influenced by proximics and attraction factors. Individuals in close 
proximity (face-to-face Job interview, task force meeting, etc.) are 
afforded the opportunity to interact and discover similar interests and 
activities. These similarities form the basis for attraction and future 
interaction (cultivation) in the development of a mentoring re1ations^l1p. 

A fourth characteristic seems to be the pervasive imporf^ce of 
the informal mentor* ,g relationship in fostering career development and 
success for the protege. For example, in Roche's [12] study of 1,250 
executives, 63X indicated they had a mentor. Furthermore, those 
executives with a mentor reported higher salaries, bonuses, total 
compensation and career satisfaction than non-mentored executives. Other 
research indicates most corporate presidents have had mentors at some 
stage of their career [•.]. As the title of one article puts it - 
"Everyone who makes it has a Mentor" [21. 

Beyond its Impact of personal and career development, mentoring 
can also impact broader organizational functions and activities. Robert*. 
[10], for example, includes sponsoring (mentoring) as one of the five key 




roles personnel must perform which are necessary for effective 
performance of the R&O function. Roberts notes that, beyond providing 
protection, coaching, and encouragement, the mentor can help establish 
the appronrldte organizational culture for effective R&O. Zey [14] 
discusses at least seven benefits to the organization Including 
enhancement of the processes of management development, management 
succession, and socialization to power. Another benefit, of particular 
Interest here. Is that of Improved organizational comnunlcatlon. 

Zey views the mentoring relationship. In part, as a means cf 
promoting comnunlcatlon between various levels of the organization In 
that the mentor and protege act as "linking pins" In the sense that 
Likert used the term [7]. These linking pins thus act as a communication 
channel by which Information can flow between two management groups at 
different levels In the hierarchy. While Likert conceived the linking 
pin relationship to be a formalized one (between members of the chain of 
conmand). Zey, 1r his research, found that proteges he Interviewed tended 
to perform many of the linking pir; funct1o?ts In their Informal mentoring 
relationships. , 

The Informal mentoring relationship also has Its risks. X:. a 
study of 3000 mentor-protege pairs, Blotnick [1] f und only 34 pairs were 
i*ble to maintain the relationship for three or more years. Moreover, 
1200 of the 3000 proteges were eventually fired by toe mentors! While 
not all studies Indicate such drastic outcomes, research on the phases of 
Informal mentoring relationships does Indicate that change, 
organizational and/or Individual, can place strain on the relationship 
which leads to conflict and separation, often with negative 
consequences. For example, as the protege gains confidence <:nd 
experience, he or she may desire more autonomy, thus pressuring the 
mentor to ""loosen the reins". Should the mentor perceive the need for 
autonomy as preiriature, or If the mentor has become too de')endent upon the 
protege, confllci develops due to Individual change. Organizational 
changes such as promotions and transfers also can leave the participants 
In a mentoring relationship feeling as 1f they were left 'holding the 

% ■ 



A question that arises at this point Is: "can some program be 
created and managed which maintains the advantages of the mentoring 
relationship while reducing or eliminating the disadvantages?" Tiie 
general response seems to be "It may or may not be worth a try" with some 
organizations Implementing formal mentoring programs and other 
Individuals polntlnr, to tiie risks of such fonnul programs [6,9]. 

Hentorinq - Lh£ Formal Approach 

One company that has launched a formalized mentoring program Is 
Federal Express Corporation. Because of rapid growth and a promotion 
from within policy, the corporat'.on has a very specialized management 
workforce without tne broad \?»t of experience needed to be the future 
leaders of the cvvany. A formalized mentoring program was Identified as 
a means of "cross-fertilizing" these manager^. 



^'V^ESPP-'-." . 



f Four years ago, one division Implemented such a program, and It 

has proven successful. In 1984, a Leadership Institute was founded In 
t which outstanding managers were selected as Instructors, or "preceptors", 

• for a twelve to fifteen month tenure. These people were targeted for 
^ participation in a revised and revamped mentoring program. 

} In the program designed for the preceptors, there are four roles 

; or functions: (1) the preceptor/protege, (2) the mentor, (3) tne natural 
boss, and (4) the HRO coordinator, who is responsible for working with 
the people involved to facilitate and track the relationships. Some of 
the critical elements of the program design are: 

1. the program is a voluntary developmental opportunity; 

2. the preceptors must select a mentor who Is from a different division 
and Is two levels above the preceptor; 

^ 3. both parties must agree to a no-fault conclusion. 

The Inclusion of these elements has been found to b»»st facilitate the 
needs of both the organization and the people Involved. 

■ ■ Thro ^h a series of self-analysis activities, the preceptor 

identifies xnree executives for consideration as mentor. After a mentor 
f candidate is interviewed by the HRO coordinator, the preceptor is told 

I. who is interested, makes a selection, and participates in an introductory 

'^^ meeting, which is facilitated by the HRO coordinator. 

; The true success of these relationships Is then a reflection of 

the time, energy and interest dedi ated xo them by the mentors and 
proteges. The HRO coordinator periodically checks on the status of the 
pairs, and the natural boss Is encouraged to review and reinforce the 

r: activities and discussions in which they are Involved 

At f^resent, the Leadership Institute has seven preceptors: six men 
.nd woman. There are two other women, outside of the Institute, who 
sought developmental support from HRO and are also involved In the 
program. Only one member of this group, a woman, 1s black; none of the 
mentors are black. All of the proteges are between 30 and 40 years old; 
the mentors range from 40 to 55. 

Mentoring and Communication 

As mentioned earlier, the mentoring relationship can have an 
; Impact on the organi^.ation beyond enhancing personal and career 

development. In particular, if the formal mentoring relationship 
discussed above is viewed as an organizational communication channel, 
there intuitively would seem to be irnpilcations for information 
dissemination, innovation, and productivity - factors which have been 
proposed as essential In post-industrial society. 

r : A conceptual framework fo.' exploring the relationships betwetn 

;•: nentoring, organizational communication ard innovation and productivity 







: i 

Is Indicated In Figure 1. First, mentoring Involves performing certain 
roles described earlier (sponsor, protector, etc.). [6] Second, 
performing these roles entails and evokes, at the same time, specific 
communication dimensions characterizing organizational communication. 
[11] Third, the combination of roles and communication dimensions 
implies the existence of an organizational network which -an help create 
organizational conditions which have been found to facilitate Innovation 
and productivity [5.8]. 


Relationship of Mentoring as a Connunl cation Channel to 
Innovation and Productivity 

Communication Factors 



Information Load 





Mentoring Roles 


Exposure & Visibility 
Challenging Assignments 

Role Model 
Acceptor & Conflrmant 

I Cownunl cat Ion Network j 


Environment for Innovation & Productivity 

Technical & Political Information 

Whollstic Thinking 

Long-range Thinking 


Support Network 

Attitudes of Optimism, Confidence & 

Risk Taking 
Reward (eficouragement) of Innovation 






An example might help. One factor commonly associated with 
Innovatlo.i Is the existence of a long-range perspective. Conceptually, 
this perspective could be conveyed to a protege by the mentor performing 

* a number of roles. By providing assignments which require long-range 
thinking, the protege Is challenged to Incorporate this perspective. 
Furthermore, the assignment might be one with exposure and visibility 

I where the protege could come Into contact with key organizational 

personnel who are known by the mentor to value and encourage this 
perspective. Whether this mentoring strategy Is effective, however, also 

J depends upon the n?ture of communications within the network. For 

* example, If the "wntor does not trust the protege, she/he Is not likely 
to provide challenging assignments which have exposure and visibility. 
If the protege rec Ives too much Information (overload) from the mentor 
and/or key personnel, learning effectiveness is likely to be decreased. 
In summary, these examples Indicate a delicate balance between mentoring 
roles and communication dynamics In creating and maintaining an 
organizational network for enhancing information dissemination, 
innovation, and productivity. 


While the for.going model proposes logical relationst.'ps between 
mentoring, communication, and Innovation and productivity, the concepts 
in the model are based largely on research results Involving Informal 
networks. Very little Information is available concerning formal 
mentoring relationships and even less Information is available concerning 
the Impact of formal mentoring programs on innovation and productivity. 
Thus, the applicability of the model to a formal setting was Investigated. 


Interviews were conducted with 11 members of mentor-protege teams 
Involved In the Federal Express program. No specific hypotheses were 
established prior to the Interviews. Rather, an Interview format was 
constructed so as to elicit personal insights Into three areas of 
Information: (1) the mentoring roles (sponsor, coach, etc.) that were 
prevalent in the relationship, (2) the communication dimensions (trust, 
gatekeeping, etc.) that characterized these roles, and (3) the Impact the 
relationship has had on personal and organizational innovation and 

An open-ended format was used to allow mentors and proteges to 
relate their insights Into these three aspects of the relationship. 
Given the exploratory nature of the research, It was felt that sample 
members should have as much freedom as possible to provide Insights which 
may not be represented In the model. At the same time, however, the 
Interview format was structured so as to require sample members to 
comment, at some time in the Interview, on the specific dimensions of the 

All interviews were conducted by one of the researchers who 
previously had established a rapport with sample members. This 
Individual had participated In the development and implementation of the 



.V^v;-..r-. ' \^ 

%\ ■>i^-tmm "v -♦ 


formal mentoring program and had Interviewed the sample members at 
various times during the life of the program. The contents of the 
Interviews were recorded and transcribed for analysis. 

The Interviews produced 121 single-spaced pages after 
transcription. Both researchers worked separately on the analysis which 
Involved an Inductive approach to generating propositions about the 
model. Each researcher generated propositions which could be Illustrated 
by quotes from the transcripts. The researchers then met and combined 


In general, the findings support the model. Numerous examples 
were found which Indicated the mentoring relationships did act as a 
comnunlcatlon channel and that information dissemination. Innovation and 
productivity were Impacted by the mentoring relationships The results 
also Indicated, however, that the relationships were not equally 
effective. Rather, certain factors seem to be needed for the formal 
relationship to achieve Its potential. 

A first proposition Is that the formal mentorlug team must go 
through developmental phases similar to those found In Informal 
relationships. This proposition evolved out of a comparison of the 
responses from newly formed teams and those from teams which had 
previously some Informal relationship. The following quote from one 
mentor, who Is Involved In both kinds of relationships. Illustrates this 

It takes a while for each of those two personalities to know 
the other one and feel comfortable. . .In the case of people I 
worked with for a longe" period of time, we had been through 
that cycle... you have to go through a whole series of subjects 
and discussions until you find out that you respect each 
other's competence. Then you go through another phase where 
the, I guess, the respect level continues, and then It finally 
gets to something that Is, I guess, trustful. 

A second proposition Is that the potential for the formal 
relationships to Immediately Impact Innovation and productivity Is 
limited by the extent to which the parties In the relationship see this 
potential. When questioned about the benefits of the relationships and 
whether It had any effects on personal or corporate Innovation and 
productivity, most of the Individuals viewed the major Impact In terms of 
long-term management development. Wheii asked questions such as, *Can you 
describe a situation or conversation you've had with your mentor/protege 
in which they helped you solve a problem or your mental 'light bulb' came 
on?", most Individuals could not give an example. 

A third proposition Is one related to the second proposition. The 
potential for the relationship to Impact Innovation and product'vlty also 




seems limited, Initially, by the cross divisional structure of the formal 
I relationship. While one of the human resources objectives of the formal 

relationship Is to provide the protege with a mentor from another 
p division, and thereby broaden the protege's corporate perspective, the 

difference in perspective Is, Initially, a barrier to be overcome. The 
barrier seems manifest In at least two ways. First, both mentors and 
proteges expressed a concern about the accountability of what the protege 
learned. For example, one mentor was concerned the protege would only be 
able to apply the experience In the mentor's division. Second, in those 
■ relationships where the mentor and protege had some prior relationship or 

i were within a common division, the common background was found to 

facilitate Innovation and productivity more so than In the 
cross-divisional relationships. This factor was particularly clear In 
the Interview when the protege Indicated her mentor (from the same 
division) was able to brainstorm Ideas with her. 


■ » 

J As exemplified by the Federal Express case, the formalization of 

the mentoring process Is possible, and has the potential for both 
Immediate and long-term benefit to the Individuals Involved and to the 
corporation. However, It seems that the chemistry that occurs In 
Informal relationships must still be established In order for the 
formalized relationship to be fully productive. While Informal 
relationships typically are founded on friendship and a mutual trust, 
formalized pairings must take some time to determine where they "stand" 
on critical Issues. 

Additionally, since they are founded on the premise of personal 
development, most formalized pairs seem to have some difficulty In seeing 
the utility of the learning that occurs beyond the relationship Itself. 
This narrow focus Is further emphasized by the barrier created by the 
matching of cross-disciplines In the pairs. In order for these 
relationships to enhance the Innovation and productivity of the 
corporation, and not just the Individuals Involved, It would seem 
appropriate for these Issues to be addressed at the Initiation of the 

When viewed from the perspective of the organization with a large 
research and development population. It would seem that a formalized 
mentoring program would serve as a valuable development tool for the 
Individuals and an investment In the future for the organization. With 
the proper pre'iaratlon and faclllfjtion of the Individuals and their 
pairings, a mentoring program could not only broaden the people Involved 
beyond their individual projects and responsibilities, but could also 
increase the probability of a smooth and progressive Integration between 
their current activities and their organization's mission for the future. 




tJ^^lkUn kr i*^' -JtV. . 


Let Avant Is a Sr. Human Resource Development Specialist at Federal 
Express Corporation, working with the Leadership Institute. In addition 
to her Interest In mentoring relationships, she Is studying thinking and 
learning styles. Robert W. Boozer Is an Assistant Professor in Memphis 
State University's Fogelman's College of Business and Economics, 
Department of Managenient. His research Interests Include stress, 
neurotlcism, mentoring, communication and Innovation. 




•lrrt*ii'**kVi\i\ -v^...--^., . " \^ 


[I] Blotnick, S., "With Friends Like These..." Savvy . October 1984, pp. 

[2] "Everyone Who Makes It Has a Mentor", Harvard Business Review . Vol. 

56, No. 4 (July - August, 1978), p. 89. 

[3] Huber, George P., "The Nature and Design of Post-Industrial 

Organizations", Management Science . Vol. 30, No. 8 (August, 1984) 
pp. 928-951. 

[4] Jennings, E., Routes to the Executive Suite . McGraw-Hill, 1971. 

[5] Kanter, Rosabeth M., The Change Masters: Innovation for 

Productivity 1n the American Corporation . Simon and Schuster, Inc., 

[6] Kram, Kathy E., Mentoring at Work: Developing Relationships In 
Organizational Life . Scott, Foresman and Co., 1984. 

[7] Likert, R. New Patterns of Management . McGraw-Hill, 1961. 

[8] Mendell, Stefanle, and Enn1s, Daniel M., "Looking at Innovation 
Strategies", Research Managemen t. Vol. 28 (May - June, 1985), pp. 

[9] Phillips-Jones, Linda, "Establishing a Formalized Mentoring 

Program", Training and Development Journal . Vol. 37 (February, 
1983), pp. 38 - 43. 

[10] Roberts, Edward 8., "RiO In the Context of the Whole Organization", 
Research and Development: Key Issues for Management . The Conference 
Board, Inc., 1983, pp. 42 - 46. 

[II] Roberts, Karlene H., and O'Reilly, Charles A., "Measuring 
Organizational Communication", Journal of Applied Psychology . Vol. 
59, No. 3 (June, 1974), pp. 321 - 326. 

[12] Roche, 6., "Much Ado About Mentors", Harvard Business Review . Vol. 

57, No. 1 (January - February, 1979), pp. 14 - 28. 

[13] Shaw, Marvin E., Group Dynamics: The Psychology of Small Group 
Behavior . McGraw-Hill, 1971. 

[14] Zey, Michael G., The Mentor Connection . Dow Jones-Irwin, 1984. 



-f^'w<». ' v^ .. . ■->»^ .^ 





s . ■- r 



William J. Murphy, Harvard Business School 


As cooperative ventures to condurt research and development become irx;reasinofy attractive, 
manaaemerrt ot these collective ur>dertaklno8 poses new challenges to executives. In selecting 
the most appropriate organizational structure and strategy the author suggests that ttie ootlective 
enterprise exeaitive must strike the best balance among three distinct but related elements. 
Eight types of cooperative R4D ventures are propoi»ed with discussion of the unique 
management tasks associated wKh each type. 


Companies are not independent entities but exist in a complex web of erteinal 
relationships with other companies and governmental institutions that range from minimal 
interaction, as exemplified by short-term contracts with buyers and sellers, to the full 
integration of mergers and acquisitions. In between these iwo extremes lie joint or 
cooperative ventures. 

Continuum of Interaction 



Contractu el 
Joint Ventura 

Joint Ventura 


Form of Cooperative Activity 

Cooperative ventures also exist along a continuum of interaction. On one hand are 
the contractual cooperative ventures which, in their simpliest form, are merely agreements 
between two or more companies regarding a specified exchange of performances. At the 

'i;vxi;niNCi pagi-: blaMC not filmed 


^i^^j^nfro«wii*a< ww 

^z-^sSf^--." .- 

other extreme arc equity cooperative ventures wh'ch provide for joint decision making 
within the context of a jointly-owned ent«prise. Neither contractual nor equity cooperative 
ventures are new. 'i Although the creation, existence and death of cooperative ventures 
have been going on for decades they have received increased attention lately as more 
corporations look to collective activity to achieve strategic objectives. In particular, there is 
growing interest in cooperative ventures to undertake research and dcvelopmenL 

The [mportance of R&D and the Need for Cooperation 

Although "technology" and "R&D" are commonly used words, it is doubtful that a 
common meaning of the terms can be implied from this frequent usage. For the purpose of 
clarity the following definitions will apply. The term technology describes the knowledge 
required for the pr«3uction and delivery of goods and services. Likewise, technological 
innovation is the process by which this loiowledge is developed and ultimately transformed 
into specific goods, products, and services. One should note that these definitions 
encompass not only technology that is directly traceable to "scientific" knowledge but also 
includes knowledge in areas that have not been or can not be classified as scientific. But, 
this only takes care of the "R" part of RScD and all too often this is a common ommission 
on the part of those discussing R&D policy and management Development is an important 
and easily overlooked part of the process that takes the new information or knowledge that 
results from research and transforms it into a form that is useable, either as a base on which 
additional knowledge can be built or as an actual product or service improvement I *^' p-^1 
This transformation process is particularly important with regard to cooperative research 
and development ventures. 

The first crucial steps in the complex process of technological innovation are 
research and development activities through which increased understanding and control of 
various "technologies" are gained. Economist Joseph Schumpetcr's concept of the proc'sss 
had three distinct segments. The first segment is invention, that initial insight that identifies 
and defines a new capability. Following invention is iimovation, the transformation of the 
capability into a form useable by society. The last phase Schumpeter referred to as 
imitation, which describes thi diffusion of technology as others copy and make 

improvements on the orignal;*'» P-^U Some authors contend that too often 
the invention phase is over emphasi^id as the critical part, whereas innovation and imitation 

(or diffusion) maybe be of equal or perhaps greater importance to society.l'^' F-*^ 

Sctiumpctgrian Notion of the K&D Process 


"" Invntlon ► Innovation ► Imitation 

•li ... 

I 76 

In that a society's or institution's limited and valuable management tLme and energy 
must be allocated among various competing activities, an examination of the relative 
importance of R&D activities, collective or otherwise, is rrquired. There is little question 
that technological innovation plays a critical role in modem society. This observation is 
particularly pertinent to the last fifty years during which industrial, social, med.val, legal, 
and organizational innovations have lead to incredible economic (and ultimately social 
progress) that has been experienced by nearly all countries and peoples. Despite its 
recognized importance to prosperity and the promise of a "better life" the innovation 
process is only a vaguely understood phenomenon. This perceived importance of 
innovation is evidenc^ by an increasing chorus of concern regarding the state ot R.feD in 
this country and others. 

Research and development activities are generally cited as having four significant 
positive effects on the "commonweal". First of all, R&D, as part of the overall innovation 
process, is credited as the primarj' driving force behind economic growth.^'* p-^*1 The 
second major contribution claimed for research and development activities to economic 
health is in improving international competitiveness. In fact, some analysti of the appai ;nt 
decline of competitiveness on the part of United States firms point to a lack of innovation as 
a possible cause.^^l R&D activities as an important aspect of national defense efforts is 
generally recognized as tlie third area of contribution. Since the United States bases much 
of its defense strategy on the concept of quality instead on quantity, any threat to the 
technological lead the U.S. enjoys in advanced weapons systems '■ se for serious 
concern. As a consequence a major portion of federally-supported R&i> ^ocs into defcns.;- 

related projects. l^- p-^^' The final area of societal contribution cited for R&D activities is 
employment. It is argued that as certain industries mature and decline new products and 
services as well as new methods of production are the only hope to replace lost jobs and 
generate new ones.H. p-98] jhis employment debate is complicated since some argue that a 
portion of R&D actually rcsiilts in the reduction of employment opportunities. 

Three forces are at play in the modem economy that make cooperative R&D 
ventures increasingly attractive. First, marketplace pressures exerted by foreign 
companies, most notably tlie Japanese which have a Iiistory of cooperative R&D activity 
orchestrated by Mffl and other governmental institutions, have caused domestic 
corporations to reexamine the prevailing "go it alone" attitude. Second, the amount of 
resources (both finanical and human) necessary to carry out many modem-day high 

This employment debate is complicated by the fact that some argue that a portion of 
R&D actually results in the I'^duction of employment opportunities. Economists, in 
recognition of the fact that technological innovation can either create or destroy jobs, refer 
to efficiency increasing innovation as either "factor biased" or "factor neutral". An 
innovation is said to be factor neutral if adoption of the innovation does not result in : 
change in the relative quantities of inputs consumed per unit of output (assuming 
constant relative prices). Conversely, an innovation is factor biased if adoption results in 
a change in the relative quantities of mputs consumed per unit of output. To give an 
example, let us say tha' a machine has been developed that can weld automobiles at a 
fraction of the cost of human welden. Industry adopt* the new machine {but dc iot 
change output because of the adoption). That innovation is said o be factor b'ascd in 
favor of capital. Such factor bias, at least in the short-term, v.'.i ause unemployrienL 
it is this type of innovation (and the supporting policies such as certain tax treatnscnts) 

that appears to generate a substantial amount of controversy. f^- P'^^' 






lecnr jlogy projects has become so vast that fewer individual companies can tackle these 
projects as independent entities.!'^) And finally, changes in the antitrust prohibitions 
against f^ooperative research and development efforts have lessened ihe legal uncertainties 
surrounding joint activities among competing companies l^> ^1 

PoTces Encouraging Cc?operative R&D 

One example of this new breed of cooperative R&D effort is the Microelectronics 
and Conr.putcr Technology Corportioii or MCC, a collection <-r t i United States computer 
and component manufacturers headquartered in Austin, "ica«s. u . 'he stated objective of 
MCC to help bring into being the fifth generation of compuierv. i ..j importance of the 
;u:hnology being developed by ;his cooperative venture is veil-recopniz^d, not only by th^ 
companies directly involved and associated industries that stand to bc-nefit uut also by 
govemmentai agencies. But common recognition of the signifi. jnce c f the technology 
docs not necessarily lead to a common approach regard- ng its development. 

Just as the MCC rer- archers will be rcquircc to develop new technology, MCC 
managers 'miU be required to oevelcp manrgemcnt systems and techniques to deal with 
the unprecedented problems facing a collaborative effort of this nature. One of the 
management tasks facing retired Admiral Bobby Ray Inman, President, CEO and Chairman 
of MCC, will be to forge a consensus among the participants regarding solutions to shared 

Management u( Cooperative R&D 

r^ I 

Managing a cooperative R&D venture differs in important ways from managing a 
single participant R&D organization. The data suggest that the tasks facing the general 
manager of a collective un&rtaking are stgnificaiiJy different from those facing the general 
manager of a single participant organization. In particular, there is evidence that in a 
cooperative venture the essential execurlv«; functions of M) establishing and maintaining a 
workable communication system among the participants, (2) securing the necessary efforts 
and resources from them, and most importantly, (3) formulating and defining an acceptable 
purpose are complicated by the presence of multiple sponsors. 



Ar u>e number of participants nses there is a corresponding increase in the need for 
negotiatjo.. among them. The multiplicity of participants in a cooperative venture places the 
general manager under increased pressure, when comF:jed \i the general manager of & 
single participart enterprise, to discover and put into opcratior methods of accommodation. 
This process of accommodation and negotiation •'ndenaken by executives of cooperativs 
ventures expresses itself in (1) the sti^tcgy- making process, (2) the organizational 
stnicture, (3) the compensation and conni systems, ir.d (4) the retaurce al'ocation 
process. As more and more companies find cooperative research and development 
attractive, the need to understand the managerial tasks associated with such coUab. rative 
ciganizations increases. 

For analytical puiposei cooperative R&D activity can be divided into three distinct 
but nonetheless interrclzted elements: (1) contribution of tlie participants, (2) creation of 
benefits, and (3) transfer of benefits to contributore. The success of a cooperative R&D 
venture depends on how well tiiesc three elements are balanced with one another. For 
example, in up a cooperative R&D venture the participants often focus heavily on 
the fint element, determining what each party will contribute to the effort This can lead to 
the adoption "*" structures and procedures that hinder the creation of tht . jught after R&D 
knowledge a. : is transfer to the contributors, the second and third elements. ThetransfT 
of technoloi;y within die context of a multi-firm cooperative venture is a particularly 
important and difficult management tisk Again using the three-clement model, various 
methods to case the difficulties of technology transfer (element three ' 'the model) can be 
obtained by the adoption of specific management practices. 

Elements ui'Cooperativp R^ n Ventures 

The reasons that propel individual companies to cooperate in a collective R&D 
effort differ according to the ability or necessity to share costs and benefits. Different 
sif-^'tions involve different potential costs and benefits to the pirticipants. Mainuiining an 
accepfablc distribution of costs and benefits among the pa-licip-.nts is one o. the 
fundamental tasks facing the general managt.- of a collective enterprise. In asscssinf 'Jic 
attractiveness of a cooperative R&D venture the individurl participant is basically concmied 
about two linkages with the collective venture. One is the individual participant'^ required 
contribution to the collective venture and the otho- is die flow of benefits from the i ollcctive 
venliire to the individual participant. The decision to join and continue in a cooperative 
R&D vt nture will be determined by the -pective parti^;ipant's assessment of the relative 
value oJ these two linkages. 

In the flow chart that follows, the required contribution from the participant in bot>. 
tangible and intangibl'- resources and assets is labeled C. The benefit expected to be 
derived from ihe cooperative a^.::vity is labeled B. As long as the perceived value of B is 
greater than th-; perceived value of C. the participant has an incentive to maintain 


— - ~^- ^ - 

. ^**. *%»;■ 


cooperation. It is the function of the joint venture executive to establish and manage a 
cooperative structure that yields t>enefils in excess of contributions. 

Referring to the three-element model of a cooperative R&D venture, the general 
manager of the collective activity can adjust the relative perceived values of B and ^, and 
thereby the value *^ the cooperative venture to the participants, by managerial actioM that 
affect any one or combination of the three elements. In odier wonis, the collective venture 
executive can increase the value of the collective activity by: (1 ) decreasing the contribution 
required, (2) improving the efficiency and effectiveness of the collective venture's ability to 
transform contributions into benefits, or (3) improving the effici'incy and effectiveness of 
mechanisms that transfer benefits back to the contributing participants. Often the nature of 
the cooperative R&D effort itself or the contractual agreement among the participants sets 
limits ' •! how much of a change in a participant's contribution can be affected by 
mar- ,■ ai action, so the general manager of a cooperative effort generally must concentrate 
on acuons that focus on the latter two cations. 

Cooperative R&D Venture Flow Chart 

Free- rider Benefits 
to Non-contributors 


The literature abounds with advice and counsel to the R&D manager, of a collective 
venture or otherwise, regarfing how to improv; the efficiency or effectiveness of th.T the 
process that transforms financial and brainpower inputs into new technology and 
m ;ntions. But, there is an important difference between inventing a new t-whrology and 
getting the technology employed in useful products and services. To emphasize this point 
conccmmg the impoitancr of the development pha?e of R&D consider the example of 
penicU m. Every schooIchUd is dutifully taught that Alexander Fleming discovered 
pemcilhn and, by rniphcation, that his "discovery" was the single most important event in 
mtroducmg pemcillin to society. The fac'.s teU another story. In 1928 Fleming discovered 
that tnc mold Pemcillium mtatum produced a substance that inhibited bacterial growth It 
wasn t untd ten years later that the subs'ance was isolated and identified by a large number 
of scientists and researchers who had spent many dozens of man-years on the effort 
Millions of dol'ars and hundreds of additional man-years were subsequently invested 
before a cbmcally useful dnife was obtained. 

As one can readily appreciate the expensive and time consu -ing porUon of 
pcmcUiin s dcvelepment was not Fleming's ba-^^c discovery (or invention in Schumpeter's 


C - c^ 


conceptual framework) but the unnoticed, unheralded, and often mundane work that took 
the discovery from a scientific capability to a useful and cost effective product As tlie 
National Academy of Sciences reported in its sunmuuy of the August 1976 Woods Hole 
Workshop: "Much of the co5t and time arc associated with the stages beyond the generation 
of the basic technology itself, specifically, with the production and marketing of new 
products made possible by new technology ."l^- n>i2-i3) 71^, importance of both aspects of 
R&D is recognized by the Organization for Economic Co-operation and Development 
(OECD) in its "Rrascati Manual" which proposed standard practices for surveys of research 
and development In die Frascati Manual, research and experimental development are 
deHned to: "comprise creative woiic undertaken on a systematic basis in order to increase 
the stock of knowledge, including knowledge of man, culture and society, and the use of 
this stock of knowledge to devise new applications".!** P''l 

Types of Cooperative Research and Development 

The motivation to seek out partners for a collective R&D undertaking differs 
according to individual circumstances, but despite the wide range of possible reasons for 
engaging in cooperative R&D venture company, motivati(Xi can be broadly grouped into 
the following eight categories. It is important that the collective venture participants as well 
as the collective venture manager be aware of which motivation categories apply to the 
situation at hand and what is the relative importance of each motivation category to the 
participants. The reason for this awareness is that different organizational structures and 
strategies are appropriate for the various motivation categories. Although eight separate 
motivation categories are suggested it is often the case that any individual participant faces a 
combination of motivations, some more imporfuit that others. Likewise, it is expected that 
no two participants in a cooperative R&D effort experience the exact same set of 
motivations in the exact same ranking of importance. These differences in motivation 
further complicate the collective venture manager's task. 

1. Cooperation as a way to attain scale economies. 

Achievement of scale economies is often an important element of an individual 
company's motivation to participate in a cooperative venture, but there is an important 
limitation to scale economies as a motivating force for corporations. If the projected 
benefits from the cooperative effort are seen as yielding significant improvements in 
competitiveness it is unclear that a corporation would want to share these benefits with 
potential competitors unless the scope of the invesmicnt was so great as to exceed the 
resources of the individual participant But because individual corporate resources vary 
from company to company the "power" of this motivating element will differ. For example, 
the \TSI project in Japan, a collective research and development effort in advanced 
semiconductor technology, changed the relative market positions of the participants. As a 
direct consequence of the cooperative venture, some companies improved their compe jdve 
position vis-a-vis other participating companies and for some their relative competitive 
strength was weakened. For the companies with more limited resources the allure of 
cooperation to attain scale effects was greater than companies more favorably situated 

A cooperative effort is attractive if the collective benefit obtained (and presumably 
distributed to tlie participants) can be obtained at less cost to the individual than "going it 





■ i 

alone". The greater the disparity between the individual cost and the collective cost relative 
to the benefit, the greater is the incentive to participate. Similarly, if the perceived benefits 
of the cooperative effort are remote or uncertain, a pooling of effort can change the "scale" 
of the project facing any single corporation and thereby change the decision on whether or 
not to join. 

To the extent that the collective enterprise is 'Jssigned to exploit scale economies the 
benefits to be derived by the individual participants from cooperation should vary accoiding 
to the degree scale economies ax attained. An individual participant should prefer 
collective effort to individual effort regarding the achievement of scale economies if the 
expected value of the perceived benefits, less required contribution, associated with 
collective action, is greater than the value of the perceived benefits, less contribution, fix)m 
individual action. Likewise, the individual participant should welcome additional 
contributor/participants so long as the value of perceived benefits, less contribution, 
without the additional member is less the the value of benefits, less cost, with ttw additional 
member. Generally, this means that addition partners will be sought until the desired 
economic size of effort is reached. 

2. Cooperation as a method to permit a more efTicient use of some limited 

This second motivation category is also related to the attainment of scale economies 
but differs from the first in important ways. First of all, it encompasses more than just 
fmancial resources. Secondly, the limited resource may or may not be scale sensitive. To 
cite an example that illustrates both points, in certain "frontier" R&D projects the resource 
constraint is not money, but people, in the form of trained scientists and researchers. This 
was the situation facing both the Japanese VLSI project regarding expertise in working 
with crystals of exotic materials and the chemical industry when it formed the Chemical 
Industry Institute of Toxicology (CUT) regarding trained toxicologists. This is currently 
the situation facing MCC regarding experts in artificial intelligence and other highly 
specialized fields. 

One would expect that cooperative R&D ventures established to permit the more 
efficient use of some limited resource to exhibit somewhat different characteristics than 
collective R&D efforts to exploit scale economies. The primary benefit flowing to the 
participatits from this type of cooperation is increased access to a limited resource. To 
accomplish this objective the collective enterprise can either increase the supply of the 
limited resource available or ration the available supply. If the supply of the limited 
resource cannot be increased to meet or exceed the demands for that resource on the part of 
the participants then the general manager of the cooperative enterprise will have to devise 
and operate a rationing system. This niakes the participants potential adversaries lor access 
to the resource in short supply and the structure and processes of the collective venture 
should reflect this situation. 



(flwPW'MfcK «ki%»i"» 


■' ^^*\ 



3. Cooperation to facilitate individual investment in the development of a 
product that is not readily owned by those investing in its creation. 


Companies are naturally reluctant to make investments if the results are difficult to 
"capture" or "own". For example, investment that yields improvements in personnel skills 
rather than an actual material product are less likely to be funded since the investing 
corporation has less o\mership or control over the investment output, in this case people. 
Because of this dilemna it makes sense for the total class of potential beneflciaries to 
cooperate in conductir.g the effort As an example wimess the formation of cooperative 
i ventures like the Sen^iconductor Research Corporation and the Council for Chemical 

Research which seek to develop qualified specialists in much needed disciplines without 
having an individual corporation run the risk of funding its competitors' training. 

Another set of examples in this category would be those investments in basic or 

fundamental research. Since this type of research often results in ideas and concepts that 

! quickly spread (and whose spread is difficult if not impossible to legally contain) 

cooperative efforts in Ms area are also attractive. One of the missions of MCC will be to 

'' I develop talent and tech nology that will be difficult for the investors to capture. 

. i 

A third group of examples in this category are those investments in which the 
ownership or control of the investment output has been curtailed. As an example, the 
Toxic Substances (Control Act [15 U.S.C. §2601 (1976)] requires that certain discoveries 
regarding toxicologic al effects of workplace chemicals be publicly disclosed so that rapid 
dissemination can laine place. A cooperative venture would be useful to counteract the 
disincentive to invest Such situations have helped spawn cooperative ventures such as the 
Chemical Industr' Institute of Toxicology and the Health Effects Institute. 

With regaru to this third type of cooperation the managerial task involves control of 
the flow of benefits to non-contributing outsiders, conunonly referred to as the free-rider 
problem. A flow of benefits to outsiders who do not help pay for producing those benefits 
can jeopardize the viability of a collective venture. If the realization cf benefits by the non- 
contributing outsider is the result of decreases in the benefit streams to the participant, the 
general manager of the collective will be compelled to either seek methods to stop the 
"leakage" of benefits or force the outsiders to become contributing participants. The 
possibility of outsiders enjoying the benefits without sharing the costs will also put 
pressure on the participants to switch status to non-contributing outsiders. As a 
consequence, one would expect the general manager of such a cooperative undertaking to 
try to structure and operate the collective venture so that participant exit is hindered. One 
would also expect that the participant contribution arangement to resemble a private taxation 

4. Cooperation as a vehicle to achieve uniformity or standardization. 

There are two methods to assure uniformity if it is clear that there is a competitive 
advantage in having complementary technology. One is to achieve a monopoly or 
dominant status and thereby dictate the standard to the marketplace. The other is to engage 
in a cooperative venture to assure compatibility or uniformity. To cite one example, in the 
telecommuncations industry various devices must be able to communicate with each other. 


^Pt ■TV- iVlfS' 1, ■%■(>.• N.-o'-.l •». ...> ,» 


5^ For many years Bell Labs and Western Electric through the power of the AT&T monopoly 

•J provided the necessary system uniformity and standardization. The uncertainty (and 

f opportunity) regarding the role these entities can perform following the antitrust suit 

[ settlement has prompted other corporations to form cooperative ventures aimed at achieving 

some degree of compatibility and stan 'ardization. In the PBX field, Sperry-Northem 

Telecom, IBM-Rohm, and Ericsson-Honeywell are all illustrative examples. Evidence of 

this motivational element can be found in cooperative ventures between computer 

manufacturers and component suppliers. One alternative would b; vertical integration, but 

the resources required and the scarcity of certain critical talent would be restrictive. 

With regard to cooperation that seeks uniformity or standardization, the collective 
enteiprise executive should not be as concerned with the benefit transfer mechanisms in that 
the benefits are not normally divided up and distributed tc the contributing participants. 
Instead the general management task will focus on the contribution arrangement and the 
nature of the uniformity or standards to be produced by the collective effcxt 

" • 5. Cooperation to conduct research or develop a product that is mandated or 

J required but does not yield a competitive advantage. 

If investment is required by legal, moral, or ethical standards, and the outcome of 
the investment will not produce competitive benefits in excess of costs, then there is a 
strong impetus to seek out other corporations under the same compulsion in an attempt to 
' pool resources and share results. Possible examples in this category are cooperative 

activities relating to pollution control or employee health and safety. This category can be 
contrasted with the third motivational category in which the investment ou^ut has value but 
that value cannot be readily captured In this category are investments in which the ou^ut 
can be owned or controlled but does not have competitive value. Companies already in a 
marketplace would be likely candidates for a cooperative venture that spreads the costs of 
some competitively valueless investment they had to make. However, there would be 
reluctance to permit potential entrants to share in the fruits of such a cooperative venture 
since even a competitively valueless investment can still serve as a barrier to entry. Only 
Jiose who have already paid the price of entry will be seen as attractive partners. 

One illustrative example is the Health Effects Institute which was established to 
explore the impact of internal combustion engine by-products on biological organisms. 
One of the major difficulties facing General Motors' management in helping to set up HEI 
was determining what each participant would contribute. TTiere was no problem regarding 
distribution of results since the research results were to be equally distributed among the 
participants. GM management feared that GM's contribution to the cooperative venture 
would be so large relative to the other participants that in essence GM would be merely 
funding research necessary to its competitors. At the outset the primary concern of the 
genei^ manager of this type of collective activity will be establishing an acceptable 
contribution arrangement. After an approved cost-sharing mechanism is devised and 
cooperation is established, the general management task should shift to one of maintaining 
participant interest and support. This latter task could be a difficult one in that participating 
company interest and support of a cooperative venture with no competitive impact is likely 
to wane. 






6. Cooperation as a method to maintain independence or credibility. 

In certain conflict situations investments that seek to examine and resolve the 
subject matter of the controversy will often be more valuable if both sides to the 
controversy make a joint investment in the examination project For example, R&D 
conducted by the automobile maken regarding the health effects of auto emissions is not 
likely to have a high level of credibility with certain environmental groups no matter how 
well the actual research is done. In such circumstances a cooperative venture, such as the 
Health Effects Institute, that includes other interested parties can be particularly attractive. 
This category would also encompass projects that are subject to participatory demands by 
groups potentially affected. 

Independence of the collective effort can facilitate cooperative R&D even though the 
participants remain fierce competitors regarding the products using the technology 
developed through cooperation. Independence and credibility can also be useful in 
attracting research talent and insulatirig reseairh projects from the short-term budgetary 
focus of the sponsoring companies, hi this type of collective venture the benefit from 
cooperation is derived not so much from tb; actual output of the collective effort as from 
the nature of its production. In other words, the significant general management tasks 
regarding cooperative enterprises in this category arc more likely to be associated with the 
method of output production rather than the benefit transfer mechanism. Cooperative R&D 
ventures of this nature often arise from adversarial situations in which two or more 
opposing "sides" are trying to produce the same or similar output, usually for the 
consumption or use of a third party. Such a collective venture requires the general manager 
to design an organizational structure that keeps the participants' antagonistic demands from 
interfering with the operation of the cooperative effort Consequently, one would expect 
cooperation so motivated to be characterized by extensive negotiation among the sponsors 
during formation and then by extensive autonomy during subsequent operation and 

7. Cooperation as a means to reduce the costs or risks associated with an 
entry or strategic movement. 

A firm contemplating a business entiy into a new endeavor or a strategic movement 
in an older one can often reduce the associated cost or risk by cooperating witfi others. The 
risk or cost reduction from cooperation occurs because of the differing circumstances of the 
participants. Reducing the risk or cost of entry or strategic movement can be accomplished 
by cooperation with firms that are already favorably positioned. The KodakMatsushita 
and the General Motors/Toyota cooperative ventures are two significant examples of this 
type of cooperation. For example, in the GM/Toyota joint venture GM wants to develop 
small car manufacturing expertise, a skill of Toyota, and Toyota wants to leam about 
automobile manufacture in Uie United States, an area of GM expertise. By cooperating, 
each partner can "purchase" skill or knowledge possessed by the other partner. Since the 
"selling" partner has already fully paid for acquiring the skill or knowledge in the first 
place, the "buying" partner should be able to purchase at a lower cost when compared to de 
novo or independent action. The difficulties appear when one of the partners acquires the 
strategically-sought skill or knowledge before the other and the incentive to continue 
cooperating disappears. 


<\.>S.9lL ,'*. - "^ \l 

With cooperative R&D ventures that seek to reduce the costs associated with breach 
of entry or mobility barriers, the benefits to one set of participants are the contributions of 
others. Companies should find this type of cooperation atd:active if entry or mobility costs 
can be reduced by cooperating with fums that have already paid the price of overcoming the 
enti7 or bility barrier. Since the objective of such collective effort is, in essence, to 
assist in tnc establishment of a closer competitor in exchange for something of value, the 
cooperation shodd be relatively unstable. The ^eral management task suuounding such 
cooperation will need to focus on the difficult issues of what to do after entry or strategic 
movement has been facilitated There is also the very real possibility that one of the 
partiiers will acquire the srategically sought skills or knowledge before the others. At that 
point the satisfied partner has the incentive to cease cooperation. 

8. Cooperation to permit risk diversification or risk sharing. 

Often a firm wishes to diversify its risk by making a larger number of smaller 
investments. By placing more although smaller bets, the riskiness of investments can be 
averaged out, thereby eliminating some of the downside risk in exchange for some of the 
upside opportunity. Mining and drilling conqMnies often engag'* '.a coopendvt exploration 
to diversify the risks involved. In some situations, such cooperation may be seen as 
insurance in which a class of "at risk" entities are protected against cataclysmic change in 
circumstance. For example, R&D among competitors may insure, v^th minimal inctividual 
investment, that no one competitor makes an independent breaktfirough that would put the 
others at competitive disadvantage. 

The benefit to the participants m cooperative R&D ventures tfiat serve a risk sharing 
or portfolio facilitating function is that cooperation permits the firms to place smaller bets 
on a larger number of investments. In tliat the participant's motivation to cooperate is to 
diversify investment risk, the relationship among the participants should not, in general, be 
antagonistic. As a consequence, the management task concerning such cooperation need 
not focus as much on establishing elaborate mechanisms to maintain participant cooperation 
as with other types of cooperative ventures. Instead, developing vjduable technology and 
establishing technology transfer mechanisms should pose more of a difficulty to the 
collective venture general manager. 


Cooperative research and development ventures arc becoming increasingly attractive 
as a part of the technological innovation nrocess. Yet, despite the advantages collective 
action can offer the participants, the difxicultles in managing such consortiums can turn 
opportunity into chaos. Confusion «i:id disappointment can be alleviated by managerial 
action aimed at balancing the three elements of a cooperative venture, participant 
contribution, benefit creation, and benefit transfer, and by matching organizational structure 
and strategy to the type of cooperative venture involved. 





[11 CurUn, Patricia S., Don I. Phillips, and Ralph L. Pctrilli, editors, R&D in thc 
Federal Budget / R&D. Indu stry. & the Economy. American Association for the 
Advancement of Science, 1978. 

[2] Francis, Philip H., Prinqples of RAD Management AMACOM, 1977. 

[3] Hill, Douglas W., Co-operative Research in Industry. Hutchinson's Scientific and 
Technical Publications, 1946. 

[4] McCuUoch, Rachel, Research and Development as a Determinant of U.S. 
International Competitiveness. National T. lanniiig Association, 1978. 

[5] National Cooperative Research Act of 1984 (Public Law 98-462), 15 U.S.C. 
§4301 (1984). 

[6] National Research Council, Technology. Trade and die U.S. Economy. National 
Academy of Science, 1978. 

[7] Note, Joint Researc h Ventures Under the Antitrust Laws. 39 George Washington 
LawReview 1112 (1971). 

[8] Organization for Economic Co-operation and Development, The Measurement of 
Scientific and Technical Activftics (The Frascati Manual), OECD, 1976. 

[9] Office of Technology Assessment, Government Involvement in the Innovation 
Process. U.S. Government Printing Office, 1978. 

[10] Rosenberg, Nathan, Perspective on Technology. Cambridge University Press, 

[1 1] Schumpeter, Joseph A., History of Economic Analysis. Oxford University Press, 

[12] Secretariat, Trends in Collective Indus trial Research. Delft, Netherlands: Six 
Countries Progranmve on .'\spects of Government Policies Toward Technological 
Innovations, November 1979. 

[ 1 3] Twiss, Brian C, Managing Technological Innovation. Longman, 1 980. 

AUTHOR BIOGRAPHY: William J. Murphy is currendy working on his doctoral dissertation at the 
Harvard Business School. After having served five y ears as an antitrust trial attomey for the Federal Tnde 
Commission in Washington, D.C., from 1974 to 1979, Mr. Muiphy earned his MBA with distincdon 
from the Harvard Business School in 1981 and subsequently received a fellowship from the Harvard 
Business School to study cooperative ventures. As a doctoral candidate the author has continued his 
research into cooperative ventures, with specific emphasis on the Microelectrmics and Computer 
Technology Corporation in Austin, Texas. The data collected by Mr. Murphy over the past three yean 
cover MCC from idea to organizadonal reality and form the core of his doctoral thesis. Mr. Murphy 
teaches at the University of Massachusetts/Boston Harbor Campu.<i and the Radcliffe Seminars Advanced 
Management Program. 






Samuel H. Nassiff 

International & External Affairs Office 

Space Station Program Office 

NASA Johnson Space Center 


Memoranda of understanding have recently been signed between the 
United States (NASA) and three international Space Station partners - 
Canada, European Space Agency (ESA),'9nd Japan. The international part- 
ners are performing parallel Phase B preliminary design studies, concur- 
r(>nt with the U.S-, on their proposed elements/systems for possible 
integration and operation with the U.S. Space Station System complex. 
During the 21-month Space Station Phase P study, a large amount of tech- 
nical Interface data will have to be transferred between the U.S. and the 
international partners. Scheduled bilateral technical coordination meet- 
ings will also be held. The coordination and large number of interfaces 
required to integrate the international requirements into the Space St?.- 
tion require a "clean" interface management organizational structure and 
operation procedures to accomplish the integration task. The interna- 
tional coordination management organizational structure, management 
tool; , and communications network are discussed including the proposed 
International elements/systcins being studied by the international 


The President, in his State of the Union Message in January 1984, 
directed NASA to develop a permanently manned Space Station within a 
decade. At the same time, he also invited friends and allies of the 
United States to join in the program in order to share its benefltr.. In 
April 1985, NASA initiated a 21-month Space Station Phase B preliminary 
design study. Shortly thereafter, government-to-government cooperative 
agreements, in the form of memoranda of understanding (MOU's), were 
signed between the U.S. and three international partners. The MOU with 
Canada was signed on April 16, 1985, with Japan on May 9, 1985, and with 
the European Space Agency on June 3, 1985. Since the international part- 
ners are conducting parallel Phase B studies with the U.S., an organiza- 
tional structure has been set up within the Space Station Program Office 
to coordinate technical and operational activities with our internatloual 


.■."V*'s3^^ -.'.-' 




partners to ensure proposed interuar lonal requirements/elements are ln*'e- 
grated into the U.S. Space Station urogram. This Is accomplished through 
the International & External Affairs Of ice in the Space Station Program 

For the first time In NASA's history, a radical departure Jr • - 
gram management has taken place from that of past spaceflight pro,;, jt- . 
that four NASA "Work Package" Level C Centers (MSFC, JSC, GSFC, .1 LeRC; 
are responsible for developing specific Space Station Program El'i&;ents/ 
Systems, llie Level B Space Station Program Office, located at JSC, will 
perform overall program management and the Systems Engineering and Inte- 
gration (SE&I) function for the Space Station Program. Because cf the 
physical locations and distances involved with t-he international partners 
and that of the NASA Centers it is evident that innovative management 
techniques and a telecommunications capability is needed for voice con- 
ferences and data transfer. During the parallel Phase B studies, large 
amounts of technical Interface data will be transferred between the U.S. 
and the international partners. Multilateral and bilateral technical 
meetings will also be held. International liaison representatives for 
each partner are located at JSC for the Phase B studies. During Phase 
C/D, it Is anticipated that U.S. liaison representatives will be located 
in Canada, Europe, and Japan. The Space Station Program will be Interna- 
tional in nature, i.e, contain International elements/modules with an 
International Crew. 


NASA has conducted a number of successful spaceflight programs 
such as Mercury, Gemini, Apollo, Apollo/SOYUZ, Sky lab, and Space Shuttle. 
Each of these programs have contributed significantly to the ability of 
man to work productively and live In space. The next logical step was to 
develop a permanently manned Space Station which would operate in low 
earth orbit. Recognizing that the U.S. Space Station is the next large 
development program, substantial International interest has been exhib- 
ited due to past and present cooperative activities with NASA such as 
foreign contributions to the Space Shuttle Program. 

Following President Reagan's invitation to U.S. allies and 
friends to participate in the Space Station Program, the NASA Administra- 
tor visited Europe, Canada, and Japan for high-level discussions on 
international participation. Subsequently, the European Space Agency 
(ESA) , Canada, and Japan have signed memoranda of understanding (MOU's) 
with NASA that provides the framev/ork of cooperation on Space Station 
during Phase B preliminary design. The main features of the MOU's are 
delineated as follows: 

o It recognizes participation in prior cooperative 

o Defines cooperation during Phase B and a basis for 
longer term cooperation through develoj^-ent and 
operation phases, 




t^.:K'i%V-\ ■- 

o Identifies principles that must be defined for station 
access, cost sharing, barter, and crew participation 
during subsequent negotiations for Phase C/D/E, 

o Provides a basic description of U.S. and International 
Space Station Program, 

o Program phasing and schedule, 

' o Respective responsibilities, 

o Management reviews/liaison relv^tionships, 

o Advanced Development Program, 

o Data exchange and rights, and 

>-[ o Financial and legal matters. 


'\ Three aspects of potential cooperation exists. The first is as a 

"user" of the Space Station v;ho essentially defines missions and utilizes 
the Station capabilities. Secondly, as a "builder" who participates over 
a long term in definition and development programs and supplies funding 
and hardware, thereby enhancing the Station capabilities. And thirdly, 
as an "operator" who would participate in a specific system operation 

* on-board the Station. The three international partners are viewed as 

"builder/operator" In the Space Station Program. 

Space Station partners must be sensitive tc U.S. concerns abo 
technology transfer, exporting Jobs, and efficient overall management 
resources. The MOU's do not authorize cooperation in the Advanced De 
opment Program area. Cooperation in the Advanced Development Program 
will be considered on a case-by-case basis and entered into only when it 
is mutuallv beneficial to both sides. Matters for future discussion and 
agreement between the partners are foreseen as follows: 

o Respective responsibilities in design, development, 
operation, and utilization, 

o Principles regarding access to all Space Station 

o Pricing policy, 

o Barter of hardware and services to offset costs, 

o Length and type of commitment to the program, 

; o Protection of proprietary information and intellectual 

^ property rights, 

o Crew participation in the Space Station, 







* I 


o Definition of appropriate technology interchanges, 

o Operational costs, and 

o Appropriate legal arrangements. 


The Space Station Pro'^.rain management structure is divided into 
three levels; Level A - loc& .d at NASA Headquarters provides policy and 
overall program direction; Level B - located at the Johnson Space Center 
provides program management, budget, and technical control; and Level C - * 

located at JSC, MSFC. GSFC. and LeRC field centers provides project man- 
agement for element definition and development. Figure 1 shows the Level 
B Program Office organization. Four line offices report to the Program 
Manager. The Systems Engineering and Integration Office cstabllshefi «nd 
manages the technical content of the Spacft Station Program in response to 
the system requirements established by Level A. The Data Management Sys- 
tems and Operations Office establishes and manages the data management 
architecture and overall flight and ground op tlons. The Customer 
Integration Office establishes customer requit undents, coordinates mission 
data base, and integration of users and their requirements. The Program 
Management Office manages the program resources to the budget and sched- 
ule guidelines provided by Le^el A. The Technical Management Information 
System Staff Office is responsible for developing technical program and 
data management needs and implementing an automated computerized network 
of distributed engineering and management data systems. The Internation- 
al and External Affairs Staff Office serves as the focal point for inter- 
facing with the international community and is responsible for technical 
and management integration of International partner's requirements and 
proposed hardware elements into the Space Station Program. This office 
also serves as che focus for policy analysis, strategic planning, and 
interfacing with congressional activities. Whits House visitors, academic " ^ 

community, and other federal agencies and departments. 

International Coordination and Manay^ement Process 

An integrated technical coordination and manhgement process has 
been established to Interface wltt. the international partners (Canada, 
ESA, and Japan) to process and manage change requests, conduct formal and 
informal meetings, and provide the framework for carrying out the objec- 
tives of the cooperative prolect as established in the MOU's. The Space 
Station management functions for internation 1 participation for each 
level if. as follows: 

Level A : Provides overall policy and program direction 

o Decisions on international elements 
o Planning for evolutionary growth 
o Negotiate MO'J's 


.t^i K'^'-i^-m.s^- 

Figure 1 - Space station program level B organization 




























Figure 2- International cooperation nfianagement framework for phase B 

NASA AdmlnltlrMer 

Spact MMIon Proarain 


PaNer A PWflt Offic* 

L*vtl C 
Projacl M«im»rt 

Cenira vi 



Canadian MInlilar of fli "< 
lor Sctanca « Tachnologii 

ESA DEraclor Ganaral 

Japanaaa MMalar ol Slala 
lof Sclanea A Tactwwiojii 

Onict ol 


A CilarnalAlfairt 


Program Coordination 


WorVing Groups 

Canadian MInlalry ol 
Sctanca A Tactmolejy 

ESA Noadauaftan 

Japanaaa Sclanca 
• Ttchneletir Agancy 



Spac* Station 
Control Bot'd 












Level B ; Program Implementation 

o Technical coordination of system 

o Bastiline and control technical data base 
o Integration of international elements into 

baseline configuration 
o Establish and control ICD's 

Level C ; Project Implementation 

o SE&I support to Level B on impact f 

international elements 
o Impact of international elements on Work 

Package Centers 
o Develop system and end items 

Figure 2 illustrates international cooperation management frame- 
work for Phase B. The interfaces and coordination can be seen between 
the U.S. management levels and counterpart international levels. The 
International Technical and Integration Panel (ITIP) , located at Level B 
and chaired by the Manager of the International and External Affairs 
Office IS the forum used for technical coordination of activities between 
NASA and the International partners. Membership on the ITIP and formal 
change control flow is shown in figures 3 and A, respectively. The Pro- 
gram Coordination Committee — co-chaired by the NASA Associate Administra- 
tor and International counterpart — is responsible for overall program 
c'lrection and coordination, the main focus is decision on functional and 
technical aspects of international participation. The membership on this 
cocmiittee is shown in figure 5. The Space Station Level B Control 
Board — chaired by the Space Station Program Manager — integrates the 
Phase B study activities into baseline Space Station configuration and 
lay the basis for initiation of preliminary design activities on all 
Space Station elements. Membership on this board is shown in figure 6. 
The structure of the Space Station Control Board is shown in figure 7. 
It consists of four main panels (Operations, Customer Intergratlon, and 
International Technical and lategratlon) and the Systems Integration 
Board. These four entities are, in turn, supported by fourteen technical 
Integration panels. 

Manageaieiit Tools 

The management tools used for Phase B international coordination 
at each management Level consists of program review, coordination commit- 
tees, technical working groups, and project liaison coordination. The 
management tools are: 

Level A 

o Bilateral Program Coordination Committee 

o Multilateral Program Review 

o International Working Groups 

o Working Group on International Cooperation 

o Liaison Oversight 








LO Ui 

':> <-> 



o z 




CO • «« 

1— t 



'oui oc 





O 3 





»•-• CO 





uiu. CO 


•— • 



ou- <c 









u. >- 






U-CO »— 




oz — i 




O _l 





Z"-" «t 





Oh- 3 





•— « 

— i<: o 






•— I 




< UI CO o 






ex: Q. -^ z 






CJO < 



«t 1- 


li 1 •» 






I—OUI • 




— 1< 


zz o >- 



<t o. 



a cr 





U. •-< 

z o 





O t/0 Li 1 




— .1- 

zz:o— •!- 



1— z 


4 UJ 00 (/> 




<C =3 


1— Zct •— 


PO »- 


Z O 


tiJCO O «-■ f— 

•-• CO 



o: (_> 

Z>- •-' _IZ 


LU •-• 




•-< t/1 h- UI UJ 




f— —I 


Q£ <&:•-• 




z < 


uj»- oc o 


O O 


— • z 


UI z o »co 


— . z 




ZUI UJ >- 

«/1 z 

u. <t 


00 —t 


•-x»-»- :e 

UJUJZUlSS 00(_> o 





UI •— • 







UJ z 


ui<t <c u) t/1 CO (/> 0) (/) <o O < 



> oc 


ZQcooooaEt3-ji<:— loz 






z:£ z: -a. •«•«•« 




_i 1— 


» » 


>-< _J 






*l— 4 


UI > 


«« > 



i/jj— t/1»-<0»— ••—<•— ••—•'-< ' 

1- •- 





>-<OI— «-«l-t-»-t— t- 


oc -J 

o z 

Q. «/i 








UJ t- 


(— »- 



-1 Z 




oc • 





UJ UJ UI CO (o CO «/i 00 CO t/> o CO 

»— 4 



O tj 


1 UI 


oc UI 

<t —I 

Kt _ 





«/) U. 

Zl u. 


- a. 



«C U. 

<: u. 

c3:<<tuiaLUiuJuiuiuiuJ oui 


1/1 oc 







•— • 

u. — 

• • 





• • 


UI >- 


• * 



=» oc 



»— • 


•-• cC 





1— I— 














Ui o 










UI (/> 








k J a^ 

-• .1 

































1 h- 




Z •-" 




•-I u. 


1— o 








•-■ K- 



— i 


etc UJ 







«— « 



O -J 








<: cj 











•— I CJ 




1- UJ 

■ u. 











Qc a 






— • z 









z >-• 



UJ <C 




3e: -j 










Ul u. 




a: z 


< z 




»— « 









o •— 



Z — llJLl 






z O 

«C «CC9 


o. «_> 


oozcoz z<t 






z oz 



_j ::3 


< »— q: o CO "-I 



-JU. z 


<C UJ 



< o 


>- rsj Ul 1— Ul z 



QC Z •-■ 

o>-<z << Z OC 






1-^ ^ »— « 

QC ►-< UJ 



> <: 


_J — • UJ UJ CO 1— OC 



O Z Q. 


0»— Z Q. =3Z C9 


O -« 

O. 3 UJ O CO •-• 




a: •-• o 




o oo ■-■ 


«k « • 






a cc ac o£. cc a 






UJ ut «t 



t- »- h- 


_j a D. 

CJ O O C_) O <_> _J 




Ul Ul Ul 




•-• U) _l 


a:ac 0£ a: q: oo 


CO •-•< 


H^ »-« »-H 

1— 1 »^ •— 1 



z z 

o OO a o o_i 


O CO o 

a. 3«-« 

CO 'Jl— 

UJO <c 

• • 


(-' u. z 

• • 






Ul >- 
> QC 


•— « 








t- 1- 




=> UJ 










Ul o 










Ul (/^ 






o >- 





00 o 



►- ■-' 



UiO (/> 

♦— • 

OQ. o 

Ui ^ 

•-too • ot 


U. 00 UJ ■£ O 


u. o 00 

a. o. 

O ' •-• • O 


UJ U. Ui 



zo u. o • 


Ui o 


O •-• O O OO •-■ Ui 




MU.O u.ou.a: u. o 


•-I z 


t— o (/)t/)(/)a) u. >-< 




•f O lo oe 00 Xoc3_i o u. 

£ Ul— -O U- 


Ui »-< 




t>0 t— 


iDfcO<C •«««2E O 


<. <C 



Ui Z •U. O • UiUlUiUi o 


CO •— » 



t— ouiu-ot. a> ooooo>-t X 





Z>-->O<:00Z _►_►-.►-■ OO H- o< 


o ►-•• 



M f- •-> 00 •-• u. u. u. u. :u: < _i ac o: 

>— "*!; 



tf>u i a£OU.u.u.ij. rsiOLtoo 

a ce u. <: "un/^oooo •>— •'o-jo 


Z '-• t/1 




— • »— 


zuiozi— ui'o 00 _j oe 


0£ Z 



<: a. occoz »-»—»-»-»—< * -a 


in O Ui 



O Z UI •-• ►-« -O O O O O •-' UI Ui 

vo CO 

UI u. X 


O O h- >~ C9 1/1 UI Ui UI Ui UI O O O oO 

•— • UI 



zz«-'XZZZ':>or)-3'^oe>-<i— H- 

UJ _l 

t— «/> _l 




(^ o 

K-i »— f Ui 



QC «C ^ H-t Qi q: o£ OS QC X u. u. 
uif-Qe:asoo»— a.xoo— 1 

r> oc 




O 1— 

►-< «c o 


UJ Z U> lO z < o <C 

— • z 




ZUiUI-J c£ Of Z Z Z Z Z O Z Z Z 

U. O 

o •-< 



-<Xf-<X UI OOOOO >-vOOO 


cC Ui 




cjuizz<ca:Q.»-<i-<>-i»-<'-iz— 1<—'— 





ZC0^O0£OOI~t-H-»-t-O»-l— »- 


>- 1— — J 



UI «< ►-< C5 o: «t «t «c «t < ■-• < *t «t 


Q < 



ZOCt-OtfUlh-l-l— t-t-l-t-t-Z 

»— « 

n >- 





t-«C Z 




00 _JO 









K- eCJ— Ui Z Qi 00 O O O O O O O =» 


CQ O U) 




OOh-OOt-O <«t <<< Ui< < — • 







Ui <: •-• 




00C2O»-'t— OO0000fc0</0«/i<0000<C 


</) 1— 


< O. V- 


<c z <-• 



«k « «i *^H* C^ QC •» M • •! *c/? «k »z 


3C O > 



o:: OC a: QC CO o O a: a: Of o: OC a: o: ui 


GL >— 4 »— 1 





h- h- 





Ul<t o 





J_ oc < 





aces z 





2xXXoOC3 0XXXX5EOX2q: 

CD^-i CJ 

UI U. 1— > 

1- ZCO 


• t 


— ■ OQ 




• • 


Ui '• 

UI >- 


• • 



f- z 

> ae 






Z X 




•— 1 



a: a: 




»— • 

Ui *— • 

o oc 





t- < 

UJ o 





^ 3: 

X Ui 






UI 00 










Figure 7 - Space station board and panel structure 










































Figure 8 - Phase B milestone schedule. 


i J^a!^ ^^^^^^ ^°l^°^^^^^"l^^^^°l^^'l^^''-l^",^ 

A O O A A 

c(o mint Du* 2 m* tM 



A A 


YilAOf t *N*ir<IS/0f'lNITIOH 












,— <A 










■ AltLlfcE 


O O O A 







A A 





'. *■-'*— S^ '■•• '' . ' 


Level B 

o Program Managers Board, as required 

o International Technical and Integration Panel 

^ o Membership on International Working Group 

i o Membership on Working Group on International 

; Cooperation 

o Resident International Project Liaison 

J Level C 

T o Membership on International Technical and 

Integration Panel 
Membership on International Working Groups 
o Membership on Working Group on International 
I o Liaison, as appropriate following SRR 

J Figure 8 shows the Space Station Program schedule with the pro- 

i gram milestones. The main tool used by the Program Office for integrat- 

:] ing the activities of Lev'el B and Level C wcrk package centers, contrac- 

I tors, and International partners is a. systems refer: °d to as "Engineering 

.1 Master Schedule" (EMS). 

In addition, an options list and options matrix is used. The 
options list specifies the proposed International elements to be examined 
* across the program and the options matrix specifies which elements are to 

'j be examined in combination with other options. Basically, the EMS is a 

system that specifies twenty major program themes which have been grouped 
'^ into three categories: requirements, configuration, and strategy. The 

themes are broken down into specific engineering study activities which 
are scheduled to support the major program milestones. An example is 
shown in figure 9. 

Technical Management Information System 

An Engineering Data Base, which is used for integrating Interna- 
" tional systems and elements into the U.S. Station, is under development. 

This data base includes basellned configuration drawings, systems/ 
subsystems schematics, system requirements, schedules and plans, and 
engineering data books. The EMS, discussed previously, will control the 
content of the Engineering Data Base. Engineering, operations, customer 
integration, program management, and international interface documenta- 
tion is also under development which requires transmittal and review by 
our international partners. Additionally, scheduled NASA/International 
Partners Technical Coordination Meetings require presentation material 
transfer, action item generation, tracking, and followup. 

>; In view of the above, and coupled with the vast distances between 

i our international partners and location of the various NASA field 

.^ centers, it is imperative that a telecommunications system and technical 

j and management information transfer system are implemented. The current 

capabilities for international communications are: 


:^"*feSfe--.-' .■ 











-o, . . 

> - o ;- o 

■ i^ Is 




J - f K. 5 i 5 


O y o 6 tt % ^ 







:: ; 





f 1 

(S <s 


















> - 




or w 












i 1 


- I 

<i? . 



1 o 








— ' -^ 




.o ^ 


V 1. 




; •« - 

^ ^. J ^ ^ 

5? § 

wJ if 

^ . "i 

r — i*j 

^^' i 

JO _i 
--. *- fv 

— >*( — 1 

K 3 a 











O Li 

m >~ 

• a 1 f „ <^ 

■ «-.!£ 5 
^8 J£? 3 










— *- 




JlVfJiSfc, a:J:i. 


Electronic Mail 

- Telemall 

Text/Image Interchange 

- Facsimile 

The Technical and Management Information System (THIS) for the 
Space Station Program Is being developed in two phases. The Internation- 
al System, for Phase I (present to March 1987) Is in an implementation 
process and availability is tfirgeted for early 1986. The Phase II acqui- 
sition process has been initiated and Initial implementation is scheduled 
for March 1987. 

The THIS is an integrated system of hardware, software, proced- 
ures and people resulting in the Information and products required to 
support the Space Station Program. It is a network of distributed engi- 
neering and management data systems for information exchange linking NASA 
field centers, contractors, and international partners. 

I Figure 10 illustrates how the TMIS will be connected to the vari- 

.) ous centers through the NASA Program Support Communication Network. 

"-j Typical TMIS architecture at a NaSA center is shown in figure 11. The 

i functional capabilities of the TMIS are: data base management, CAD/CAE/ 

CAM, models/analysis tools, documents management, scheduling, planning, 
'-'. resources, electronic mall, and office automation. 


A brief overview and description is given of the proposed inter- 
national elements/modules being studied by the international partners 
t during Phase B. 

Canadian Integrated Servicing and Test Facility 

Figure 12 shows the ISTF attached to the Space Station. The ISTF 
" truss structure with its accommodations occupies: a volume of 

approximately 75 ft. x 60 ft. x 20 ft. This volume contains the 

following static 


o Positioning systems for payload servicing 
o OMV Hangers 
o ORU Pallets 
o Mobile Base 
o RSS Parking Fixture 
I; o Robotic Test Bed Accommodations. 


': The dynamic elements like the RSS, the SSMRMS, thfe payloads and 

J large space construction facilities will protrude beyond this volume 

li during operations en the ISTF. The ISTF centralizes some servicing 

functions on the Station such as payload/spacecraft serving, Integration, 
[ and checkout; OMV iiervicing and accommodations; proximity operations with 

tools and EVA work station; robotics test bed, etc. 

; 101 


^: .v>r/j'>^VVv''"'A. 

Figure 10 - Technical management and information system 




TMIt I »"C 




Figure 1 1 - Typical TMIS architecture at a NASA Center 



















Mfeiu i&Ai. ^'i. .-.-•. L "■ - ^_- y ^ 


Figure 12.- Canadian InteQrated Servicing and Tisst Facility (ISTF) concept 


(BTATHMI llil«»m«TOW) 



l ymi M I MT 






RoaoTic SERvicen ertrtH 

Figure 13- Robotic servicer system concept. 








-.-^S-'--.-' . 



The growth philosophy for the ISTF allows accoomodaclon ot evolv- 
ing technologies and evolving requirements. Sone major areas of growth 
are forseen in replacement of modified SRMS by advanced SSRMS, incorpora- 
tion of artificial Intelligence technologies and sensors into ISTF 
robotic systems for supervisory mode of control, and two HRHS operation 
simultaneously from two different control stations. 

A concept for the Robotic Servicer System is show in figure 13. 
It is comprised of a positioning arm with end effector, dexterous arm, 
toolrack, latching location, and passive grapple fixture. This system 
may be used on the ISTF foj automated servicing operations. 

Canadian Remote Sensing (RADARSAT) 

RADAR&^AT is a free-flying platform which operates in a sun- 
synchronous orbit at lOOOKM, and 99* inclination. The Earth Observation 
Satellite includes a synthetic aperture radar, a micro-wave 
scatterometer, and advanced high resolution radiometer and an optical 
sensor. RADARSAT stowed dimensions are 23 ft. long and 14.3 f'c. in 
diameter. Tip-to-tip dei^loyed solar array is 137 ft. Figure 1^ shows 
the fully deployed configuration. 

ESA - Columbus Preparatory Program 

The objective of the Space Station Columbus Program, as adopted 
by the eleven members of the European Space Agency (ESA) long-term space 
plan on January 31, 1985, is to develop the set of elements shown In fig- 
ures 15 and 16. The Columbus Program is based on previous experience 
acquired In Europe with Space Lab. The elements include: 

o A pressurized manned laboratory module whirh will be 
used as a life science and/or materials laboratory 
while attached to the Space Station. 

o Unmanned platforms for co-orbit and polar orbit 

o Unmanned service vehicle to support platform 

o Resource module to support the pressurized module 
fiee-flying man-tended option. 

The pressurized laboratory will require interfaces with the Space 
Station for power, thermal, ard coiwnunlcatlon services. It can be con- 
figured to optimize user requirements and desired mlsi>ion and pay load 

Platform missions will serve different objectives and needs such 
as material and fluid physics, life and space sciences payloads requiring 
micro-gravity and payloads for Earth observations, Stellar, and Sun 
pointing. Servicing vehicle missions (unmanned) will be utilized to per- 
form at different orbits (1000 KM/28. 5" or 700 KM/98''), operations such 


^-^ss^ ■-'.-' 

















'Vl^_»?^/-J>,3L.'' . -^*^' 


■'■' i 


as refueling or exchange of standard ORU's, visual Inspections by camera, 

Japanese Experiment Module (JEM) 

The JEM is a pressurized multi-purpnse experiment module approxi- 
mately 33 ft. long and 14.5 ft. In diameter. Figure 15 shows the IOC 
Configuratxon which consists of the pressurized module, experiment logis- 
tics module, fired manipulator, air lock, and exposed work deck. Growth 
projections shown in figure 16 Include addition of another exposed work 
deck, structural mast, OMV hanger, movable manipula''or, man-tended free- 
flyer, teleoperator, and associated service facilities. Some of the 
functions to be supplied by the Space Station to the JEM are: (1) pri- 
mary power supply and heat rejection, (2) data relay to and from ground, 
(j) primary air supply, and (4) accommodations for crew assigned co the 
JSM system. 

The crew In the pressurized module can operate a wide range of 
missions such as material processing, life sciences, space medicine, etc. 
The exposed work deck is used for accommodating high energy cosmic ray 
experiment, space robotic, liquid propellanr handling, material science, 
commercial space processing, etc. The experiment logistics module which 
is 18 ft. long and 14 ft. in diameter consists of pressurized and unpres- 
surlzed sections. The pressurized section can accommodate up to two 
crewmen and can serve as a safe haven. The module stores and transports 
experiment specimens, experiment gases, spare parts, special experiment 
equipments, 9tc. The fixed manipulator is used for servicing equipment 
on the exposed deck and manipulates or changes out co&iponents, experiment 
samples, etc. The airlock located between the pressurized module and 
exposed deck is used to transport equipment and samples with the aid of 
the fixed manipulator. 




- 1 


Figure 15- Space statron international reference configurations (Initial) 





Figure 16- Space station international reference configurations (Growth). 







- -J 

t«^A^^vt..x:-^^~^-^ ^ ■" (i 


Mr. Naulff received a B.S. in Aeronautical Engineering from the 
Unlvarsity of Florida and was eaployed at General Dynamics (Convalr), 
Fort Worth for 6 1/2 years in the Airplane Stability and Control Group. 
He Joined NASA JSC in 1963 and has been involved in the Gemini. Apollo, 
Apollo/Soyus, Spacelab, and Space Shuttle Programs in the areas of 
simulation design/training, spacecraft design, advanced missions studies, 
and engineering project management as related to spacecraft design and 
miaslon requirements. He is currently Manager of International Projects 
in the International and External Affairs Office, Space Station Program 
Office and is responsible for technical interface, management integration 
and coordination of international requirements end hardware elements into 
the Space Station Program. Mr. Nasslff is an associate fellow in the 
AIAA, and registered Professional Engineer in Texas. 








Albert H. Holland 
Universities Space Research Association 


Th'; iieed for close 'nteHependence between large numbers of diverse 
and specialized work groups makes the Space Program extremely vulnerable to 
Igss of productivity at organizational Interfacai, Trends within the program 
also suggest that the number and diversity of Interfaces will grow in the near 
term. Continued maintenance of R&D excellence will require that interface 
performance Issues be Included in any future productivity Improvement effort. 

I The types and characteristics of organizational Interfaces are briefly 

.< presented, followed by a review of factors which impact their productivity. 

"f Approaches to assessing and improving Interface effectiveness are also 



In order to accomplish its objectives, the United States Space Program 
relies upon the contributions of a wide variety of engineering, scientific, 
technical and support personnel representing a large number of diverse 
organizations. This is necessary, of course, to create a program structure of 
sufficient technical power and flexibility to fulfill its mandate. Yet 
achieving the effective Integration and coordination of so many specialized 
work groups is an enormously challenging task, and there are indications that 
this task will become increasingly complex in the near future. 

Several trends within the space program suggest that the number and 
diversity of work Interfaces will be growing rapidly. First, the sheer number 
of concurrently operating space transportation systems and facilities is 
increasing. An example of this is the development and operation of the space 
station, which will be maintained and serviced by the space shuttle. Although 
shuttle operations are now being consolidated under ». single contractor, NASA 
nevertheless continues overall supervisory activities in that program in 
addition to space station management. Second, the number and diversity of 
participants is expanding. The development and use of the space station will 
be a multinational, multidisclplinary initiative surpassing any previous 
endeavor in terms of interface management. The station will also encourage 
private sector users in pursuit of their own proprietary R&D ventures. 
Third, the indentification of opportunities for commercial profit will be 



accompanied by concerted pressure from industrial consumers to expand and 
diversify space facilities, transportation and services. To accomodate the 
broad demand, existing arrangements of critical interfaces will be pressured 
to multiply In number and to reconfigure more frequently. Finally, NASA is 
actively encouraging small businesses to enter the contractor ranks. This 
not only increases the number of contractor participants but the range of 
specialization, internal resources and experience as well. 

Effective integration will become an increasingly complex task for 
management at all levels. Indeed, the emergent skill of leadership might be 
the ability to manage work unit boundaries and regulate cross-unit 
transactions [10]. In order to do this effectively, something must be known 
about the characteristics of Interfaces, factors affecting their performance, 
and approaches to assessment and improvement. 


' Interactions and transactions between work groups flow from the 

i requirements of task accomplishment which necessitate some form of linkage or 

interface between groups. For our purposes, three basic types of interfaces 
j are of particular interest: (1) the horizontal Interface between work groups 
J within an organization, (2) the vertical Interface between authority levels, 
' and, (3) the lateral Interface between organizations. The three types can be 

considered similar in structure, with representatives from their respective 

groups interacting within the interface. Activity occurring within the 
! interface is also Influenced by the larger arganization or environment in 
'• which it is embedded. It is important to remember that the Interface can be 

viewed as a work subsystem in its own right, with its own boundary. Internal 

dynamics and degree of structure [5]. 

Representatives that populate the interface serve to Import and export 
I information and technology that is required to solve problems and reduce 

ambiguities [15]. Sometimes referred to as a boundary spanner, a considerable 
amount of research has been conducted on the power [18, 19, 23], role stress 
[1, 13] V job satisfaction [14, 15], and organizational impact [8, 19] of the 
representative occupying the Interorganizational interface. Somewhat less 
attention has been given to the activities of boundary spanners within 
vertical and lateral intraorganizational interfaces, although many of the 
activities are the same. 

Based upon the work of Miles, Brown and Schwab [7, 20], boundary- 
spanning activities may be classified into eight general categories as 

Linking : Establishing and maintaining relationships with 

representatives of other key groups; 
Importing : Acquiring task-related information, technology and 
.; resources; 

J Exporting : Distributing task-related information, technology and 

Gatekeeping : Selectively communicating information gathered at the 

fntirface back to home decision makers; 
Representing : Selectively communicating information about the home 
group to other representatives for the purpose of shaping opinions, 
behaviors, and outcomes; 




Protecting : Thwarting external pressures and influence attempts 
which might otherwise disrupt home group operations; 

Scanning ; Searching for and Identifying emerging trends or events 
which might represent threats or opportunities for the home group; 

Monitoring: Tracking environmental trends or events which have been 
identified as probable threats or opportunities for the home group. 

Brown and Schwab's study [7] of twenty-four electronics firms showed 
that the frequency of the various boundary-spanning activities varies 
according to the representatives' functional areas. For example, enginetring 
representatives tended to engage In more monitoring than manufacturing 
representatives. Furthermore, activity frequencies varied by subgroups within 
functional areas; section enginesrs engaged In significantly more linking 
activities and fewer scanning activities than did project engineers. These 
findings are congruent with those of other writers [5, 17] who contend that 
the division of labor inevitably fosters a variety of fundamental differences 
between work groups, including differences In goals, priorities, time 
horizons, formality of structure, and Interpersonal orientations. These basic 
differences in behavior and perspectives are most readily recognized when they 
are Juxtaposed as they are at the organizational interfaces. 


Factors which Influence productivity at Interfaces are those which 
facilitate or impair the occurrence of necessary transactions. Often these 
factors are not readily detectable. Drawing on the concepts and research of 
previous authors [3, 4, 5, 6, 7, 9, 11, 12 ,22, 25], nine general factors can 
be identified which affect interface productivity. These are outlined in 
Table 1 and discussed below. 

Essentiality : Within any given work system, subunlts contribute in 
various degrees to the accomplishment of overall system objectives. Some work 
groups are more critical to the task at hand than others. This is a 
by-product of the subunlts' relative centrality or position in the work flow. 
Furthermore, subunlts within a given work system vary In degree of 
interdependence with one another. The extent to which a work group views 
another as essential to its task accomplishment influences activity at the 
Interface [17]. For example, a representative of one unit may put 
considerably more time and energy into interface maintenance and activities 
than his counterpart within the same Interface. Essentiality Inevitably 
creates formal and informal subunlt power differences which by themselves can 
provide a basis for potentially disabling interface dynamics. 

Structure : The extent to which an interface is organized affects ♦^'^e 
degree of conflict within that Interface and the representatives' ^iMty to 
effectively conduct transactions [5]. The degree to which representatives 
allow information, resources and people to enter and leave the interface is 
one element of Interface organization. A highly permeable interface, or one 
that is readily open to disruptive inputs or losses of critical resources, is 
said to be underorganized [5]. An Impermeable or overorganized interface • 
boundary is one that is relatively closed to Important inputs or outputs of 
information/resources and one which rigidly maintains a fixed composition of 
people and data sources. 


■ *j'*5^ " 

^.^^?:'C^^^^- ^^^-., --^ '■"'•' U 

other elements of interface structure are clarity of representatives' 
roles, clarity of authority, and effective procedures governing formal 
Interactions. On the whole, a balanced, clearly defined, yet flexible 
structure offers the best support for interface activities [5]. 

tjiepresentatives : A person's effectiveness in the boundary-spanning role 
Is related to the choice and frequency of activities which he pursues. Brown 
and Schwab [7] suggest that representatives should be coached In which 
activities to engage. Their work Indicates that overall congruence between a 
representative's boundary-spanning activities, his Job position, and Interface 
essentially might be related to effectiveness. Certainly, neglect of critical 
boundary- spanning activities, such as scanning or Importing, could adversely 
Impact his home work qroup. 

Context: Events within the interface are Influenced by events In the 
TmineHTate and larger environment. Management within the representative's work 
unit control formal and informal incentives which affect his Interface 
activity. Representatives also frequently recruit allies from their home work 
groups and from pow-ful third parties in the larger environment [5]. 

History ; Interactions within the interface are shaped by Interface events 
which nave occurred in the past. The relationship between two work groups may 
be relatively young or old, however expectations based upon past events are 
still carried forward by representatives. Interface behavior may be shaped by 
broad stereotypes and folklore circulated within the home groups or by 
specific events experienced by the representatives. Souder [22] describes 
situations of distrust wi-ich began as individual personality conflicts and 
later became institutionalized at the departmental level. 

Factors Affecting Interface Productivity 

1. Essentiality 

. Work group central ity differences 
. Perceptions of mutual critical ity 
. Formal and informal power differences 

2. Structure 

.Boundary permeability 

. Definition of roles 

. Definition of authority 

. Effective rules and procedures 

3. Representatives 

. Congruence between activities and job position 

. Congruence between activities and interface critical ity 

4. Context 

. Control of incentives 

. Activities of third parties 

5. History 

. Specific events 

. Stereotypes and folklore 



6. Communication 

. Task-related data 

. Interface maintenance data 

. Quality and flow of Information 

7 . Norms 

. Behavior within the Interface 
. Ability to self -examine 

8. Resources 

. Delegation of authority to representatives 
. Strategic ideas 

9. Goals 

. Definition of goals 

. Differences in priorities 

. Commitment to goals and priorities 

Comtnunication : The exchange of important Information Is a key function of 
the interface. Representatives must Import and export Information of two 
types: (1) data related to tasks at hand (e.g., task requirements, 
coordination, milestones, problems) and (2) data required for maintenance of 
an effective working relationship (e*?-* Information about the interface 
itself or the nature of the transactions within it). Information exchanged 
within interfaces can either contribute to or distort the mutual understanding 
of functioning, abilities and resources across parties [9]. The usefulness of 
such exchanges further depends upon the extent to which the representatives 
are connected to internal decision makers, representatives' selection of what 
information to transmit, and information timeliness. 

Norms : Norms which regulate behavior between work groups act to support 
or inhibit interface productivity to varying degrees. For example, 
representatives who are able to openly discuss interface maintenance issues 
will more likely be able to adapt the interface to unexpected work 
contingencies. However, the ability to engage in self-examination is 
extremely difficult in an interface which is constrained by norms suppressing 
such discussion. Argyris [3,4] emphasizes the Importance of being able to 
question the fundamental norms and assumptions which govern our work behavior, 
and he convincingly describes the negative outcomes that result from not 
developing that ability. 

Resources : The extent to which work groups delegate appropriate authority 
to their representatives is important to representatives' actions. If a 
boundary spanner has insufficient power to make decision: of a tactical nature 
at the interface, then his ability to buffer internal decision makers will be 
compromised, and top management will be swamped with minor details [26]. 
Excessive representative authority takes management out of the decision loop 
and results In decisions being made without benerit of the larger picture. In 
addition, the home group is an excellent source of id^as concerning 
negotiating and Influence strategies, as well as a sounding board for 
planned initiatives. 

Goals : In order for work to proceed, specific and attainable subgoals are 
negotiated within the interface. The extent to which these subgoals are 
clearly defined and are congruent with one another is associated with the 
degree of conflict between representatives [5]. Since most Interfaces manage 



- - - -Tfir-V-, 



multiple goals corcurrently, significant differences In priorities assigned to 
mutual goals would likewise Impact transactions. Finally, the extent to which 
goals and priorities are accepted by all parties Influences the extent of 
commitment to those goals. 

Many of the factors underlying interface productivity affect 
Interfaces in overt, readily Identifiable ways. However others, such as 
context and structure, act In a subtle nanner upon elements and interactions. 
Although the actions and Interaction of the factors themselves may not be 
readily apparent, their effects generally are more discernable. Determining 
the configuration and extent of these effects permits us to identify 
opportunities for improving interface productivity. 


Although some underlying factors are more readily observable than 
others, one approach to evaluating interface productivity relies on estimating 
the underlying factors by assessing discernable effects. Some of these 
effects, symptomatic of productivity loss, are shown in Table 2. 


. Hostility 

. Extreme stereotyping 

. Severe Information distortion 

. Distrust 

. Mutual avoidance 

. Excessive competition or collaboration 

. Bilateral self-serving manipulations 

. Disruptive turnover of representatives 

. Concurrent use of redundant Interfaces 

. Little cross-party Involvement 

. Poor mutual understanding of party functioning, abilities 

and resources 
. Inflexible roles, rules and procedures 

. Inability to discuss issues pertaining to the interface itself 
.Task expectations not voiced 
. Unclear roles or points -of -contact 

. Reluctance to utilize other party expertise in project planning 
. Overt and covert task sabotage 
. Excessive agreemeent 

. Avoidance of sensitive but relevant issues 
. Appeasement 

. Suppression of disagreement 
. Decision by default or "rubber stamping" 
. Chronic recurrence of problems once thought solved 

* Based in part upon [2, 5, 6, 7, 8, 9, 22]. 




t- s 

Techniques used to evaluate R&D productivity within work groups can 
be applied to Interface assessment. This author agrees with others [21] who 
favor serni -quantitative measurement techniques (e.g., rating scales) over 
highly quantitative (e.g., ratio) or highly qualitative (e.g., antecdotal/ 
Intuitive) appro&ches. Pappas and Remer [21] suggest using peer ratings In 
which R&O project personnel rate each other In terms of productivity. 

Rather than enter the arena of individual performance appraisal 
however, ratings of the interface might incorporate the advantages of 
semi- quantitative data without the problems of peer ratings. Interface 
Incumbents could complete a survey containing Likert-type scales, rating 
characteristics of the Interface along relevant factors. This approach offers 
several advantages: 

(1) It focusses Incumbents' attention on a single subject of mutual 

Interest: the Interface; 

(2) Ratings are of salient interface characteristics, rather than of 

each other, providing a superordlnate goal instead of a source of 

(3) As a self-evaluation, the technique is mobilizing and motivating. 

All interface incumbents participate, and the data are "owned 
by the participants; 

(4) Results provide an issue-oriented focus, around which construc- 

tive dialogue can occur; 

(5) Estimates of interface functioning can be made periodically, 

providing participants with an opportunity to make comparisons; 

(6) The method is easily embedded Into a wide variety of Improvement 

programs and approaches. 

The basis for individual Items might be the underlying factors 
Influencing productivity, the ability to effectively conduct boundary- spanning 
activities, and/or the presence of positive and negative effects. Such an 
approach would indicate not only the overall health of the Interface, but 
would also direct Improvement efforts along specific lines. The Management 
Analysis Office at Johnson Space Center is currently considering utilizins a 
workshop format in which key interface managers would complete ratings of'this 
sort as a method of promoting awareness and discussion of Interface Issues. 

A sample of the types of changes that might be made to Improve 
interface productivity are shown in Table 3. The particular interventions 
selected for use in any given situation depend, of course, upon the 
configuration of reported effects. 


1. Fractionate Issues to reduce their size. 

2. Increase believable communications between representatives. 

3. Redefine mix of personnel and resources at the Interface. 

4. Clarify incentives for collaboration. 


-•'^~-*^- •••. 

5. Generate credible information and discussion regarding the 
interface Itself. 

6. Recruit third parties to regulate amount of conflict. 

7. Resolve non-controversial issues first. 

8. Increase or decrease buffering. 

9. Train personnel In interface factors and boundary- spanning 

10. Increase ambassadorship, cross-p^rty visibility and 

11. Systematize cross-party Job rotation, transients or visitation. 

12. Establish a decision authority charter. 

13. Legitimize the interface to work group incumbents ^nd third 

14. Improve communication channels between representatives and 
decision makers. 

15. Conduct a mutual check of goal priorities. 

* Based in part upon Brown [5,6] and Souder [22]. 

. Since organizations and their interfaces are dynamic in nature, 

] specific interventions must accomodate shifting factors and effects. What 

'i was appropriate last year may be inappropriate today. This emphasizes the 
importance of making specific changes within- the structure of a flexible and 
on -going assessment/improvement process . Constructive changes made 
without the supporting framework of such a process are likely to be 


As the space program enters a new era of commercialization, 
competition, and global involvement, management will be required to commit 
increased levels of effort to interface productivity. It is time to 
incorporate interface issues into existing and planned productivity 
improvement programs and research. Although boundary- spanning activities 
have been central to much research [12, 14, 16, 18, 19], little attention has 
been given to factors Impacting their effectiveness or to the relationship 
between interface functioning and productivity of the larger work group. 

There are numerous approaches to productivity measurement and 
Improvement, however managers ara in need of tools and processes which meet 
significant constraints on ,heir time, manpower and funds. Specifically, 
needed are tools and processes which: (1) minimize disruption of work group 
operations, (2) maxiniize "user-friendly" techniques (e.g., checklists), 
(3) maximize participant ownership, (4) maximize organizational self- 
evaluation and self- improvement; (5) focus only upon issues relevant to and 
under the control of the participating organization, and (6) are capable of 

- • 

j Interface performance is only one element to consider when 

:* examining the productivity of a work system. However, in an effort as hetero- 

1 genous and Interdependent as space work, its inclusion Is essential. 





Albert W. Holland, Ph.D. Is i Visiting Scientist at the NASA Johnson Space 
Center under the auspices of Universities Space Research Association. A 
licensed Industrial/organizational psychologist, Or. Holland Is presently 
working on a variety or projects and applied research In support of ground- 
based and space station productivity. 


[1] Adams, S., "The Structure and Dynamics of Behavior In Organizational 
Boundary Roles", In K. Dunnette, Handbook of Industrial and 
Organizational Psycholo gy, Rand NcNally PuETlshIng, 1976, pp. 

[2] Alderfer, C.P., "Consulting to Underbounded Systeais", In C.P. Alderfer 
and C. Cooper, Advances In Experiential Social Process . Vol. 2, 
John Wiley and Sons, 19757 

[3] Argyrls. C. Intervention Theory and Method ; A Behavioral Science 
View , Addl son -Wesley Pub! Ishing Company, 1970 ." 

[4] Argyrls, C, Increasing Leadership Effectiveness . John Wiley and 
Sons, 1976. 

[5] Brown, L.D., Managing Conflict v\t Organizational Interfaces . 
Addl son-Wesley Publishing Company, 1983. 

[6] Brown, L.D., "Managing Conflict Among Groups", In Organizational 
Psychology: Readings on Human Behavior in Organizations , 4th~i3. , 
Prentice-Hal 1 , Inc., l554 . 

[7] Brown, W.B., and Schwab, R.C., "Boundary-Spanning Activities in 
Electronics Firms", IEEE Transactions on Engineering Management, 
Vol. EM-31, No. 3, (August, 1984), pp. 105-111. 

[8] Callahan, R., and Salipante, P., "Boundary-Spanning Units: 

Organizational Implications for the Management of Innovation", 
Human Resource Management . Vol. 18, No. 1, (Spring, 1979), 
pp. 26-31. 

[9] Faas, F.A.M.J., 'How to Solve Communications Problems on the R and D 
Interface", JournaJ of M anagement Studies , Vol. 22, No. 1, 
(January, 1955y7^pp.~§3-102. 

[10] Gllmore, T.N., "Leadership and Boundary Management", J ournal cf 
Ap plied Behavioral Science . Vol. 18, No. 3, (1982), pp. 343-35?. 





[11] Hickson, D., Hinlngs, C, Lee, C, and Schneck, R., "A Strategic 
Contingencies' Theory of Intraorganlzatlonal Power", Administrative 
Science Quarterly . Vol. 16, (1971), pp. 216-229. 

[12] Jemison, D.B,, "The Importance of Boundary Spanning Roles in Strateg'- 
Oeclslon-Maklr.g", Journal of Hanagemen i Studies . Vol. 21, No. 2, 
(April, 1984), pp. 131-152. 

[13] Kahn, R.L., Wolfe, D.N., Quinn, R.P., Snoek, J.D., rnd Rosenthal, R.A., 
Orqanizational Stre ss: Studies In Role Confllce and Ambiguity . John 
Wiley and Sons, 1964. 

[14] Keller, R.T., "Boundary-Spanning Activity, Role bynaniirs, and Job 
Satisfaction: A Longitudinal Study", Journal of Business Research , 
Vol. 6, (May, 1978), pp. W-HB. 

[15] Keller, R.T., and Holland, W.E., "Boundary-Spanning Roies 1': a Research 
and Development Organization: An Dapirical Investigation", Ac ademy 
of Management Journal . Vol. 18, Ho. 2, (June, 1975), pp. 388 --93. 

[16] Kennedy, C.R., "The External Environment-Strategic Planning Interface: 
U.S. Multinational Corporate Practices in the 1980*s*, Journal of 
International Business Studies . Vol. 15, No. 2, (Fall, ^84), 
pp. 99-108. 

[17] Lawrence, P.R., and Lorsch, J.W., Organizatio n and Environment . 
Harvard Business School, 1967. 

[18] Leifer, R., and Delbecq, A., "Organizational/Environmental Interchange: 
A Model of Boundary-Spanning Activity", Academy of Management 
Review, Vol. 3, No. 1, (January, 1978), pp. 40-50. 

[19] Leifer, R, , and Huber, G.P., "Relations Among Perceived Environmental 
Uncertainty, Organization Structure, and Boundary-Spanrjlng Behavior", 
Administrative Science Quarterly , Vol. 22, No. 2, (June, 1977), 
pp. 235-247. 

[20] Miles, R.H., Macro Orqanizat ional Behavior . Goodyear Publishing 
Company, 1980. 

[21] Pappas, R.A., and Remer, D.S., "Measuring R and D Productivity", 
Research Management , Vol. 28, No. 3, (May, 1985), pp. 15-22. 

[22] Souder, W.E., "Promoting an Effective R and D/Marketing Interface", 
Researc h Management , Vol. 23, No. 4, (July, 1980), pp. 10-15. 

[23] Spekman, R.E., "Influence "nd Information: An Exploratory 

Investigation of the Boundary Role Person's Basis of Power", 

Academy of Management Journa l. Vol. 22, No. 1, (March, 1979), 
pp. 104-TT7. 



' \1 

[24] Strauss, G., "Tactics of Lateral Relationship: The Purchasing 
Agent", Administrative Science Quarterly . Vol. 7, No. 2, (1962), 
pp. 161-186. 

[25] Thomas, K., "Conflict and Conflict Management", In M.D. Dunnette, 

r Handbook of Industrial and Organizational Psychology . Rand 

\ McNally PuEHshing, ISJ^Tpp. Wi'WS. 

[26] Thompson, J., Organizations in Action , McGraw-Hill Publishing 

^ Company, 1967. 










Cynthia A. Lengnick-Hall , Purdue University 
Donald C. King, Purdue University 


Research and development units are faced with the challenging 
objective of being cost effective wl.ile developing high quality, innova- 
tive products. Advanced technology is only part of the solution. It is 
increasingly clear that organization structi'res and managerial processes 
must also be designed and structured to c&et the dual objectives of 
quality and efficiency. This paper presents the results, of an empirical 
case analysis of a large R&D division which is attempting to meet this 


Many firms in both the public and the private sectors of the 
economy view the process of generating and implementing creative designs 
and product advances as critical to their success. Competitive challenges 
for the internal resources of the firm and in the larger external market- 
place have put pressure on research and development units to be simulta- 
neously creative and cost effective. Since the Innovation process has 
traditionally been viewed as costly and inefficient, the dual demands of 
creativity and efficiency have not been easily met. 

In other functional areas such as manufacturing and distribution, 
automation has been a significant tool for reducing the costs of opera- 
tion. However, until recently automation implied reduced flexibility, 
loss of creative capacity, and high overhead making it incompatible with 
the R&D environment. The assumptions which were Inherent in the early 
developments in automation, are quite similar to the assumptions which 
guide mass production. Assembly lines, numerically controlled machines, 
and machines in which the hardware is programmed are based on the same 
concepts of standardization, specialization, routinization and simplifica- 
tion which form the basis of efficient and effective mass production oper- 
ations. Research, development, and batch production, in contrast, require 
a substantially more sophisticated technology. Machines which are, at 
minimum, programmable, and more often contain sensing and self -pro- 
gramming features are needed to effectively automate development and 
small batch operations. 




The use of computer-assistend design and manufacturing systems for 
research and development activities and for small batch production, has 
been offered as a potential solution to the difficult problem of generat- 
ing creative designs while maintaining a bottom line. Computer-assisted 
design and manufacturing techniques and computer-assited communication 
systems help to reconcile the need for flexibility with a concern for 
efficient operations. However the design of state-of-the-art technical 
systems is not sufficient, by itself, to lower costs and simultaneously 
enhance innovation. Problems in the effective use of advanced technology 
more often stevn from poor implementation or from managerial uncertainty 
regarding how to capitalize on the flexibility provided than from techni- 
cal design deficiencies. In fact, a number of research studies (e.g., 5, 
3, 8) indicate that non-technical organizational factors were the crucial 
barriers to effective and efficient innovation. Thus, the design of a 
state-of-the-art research and development effort must include the design 
of appropriate organization structures and processes. 

While the notion that organization structures and processes are 
important factors in achieving organization effectiveness is not new, 
this concept has often been overhsadowed by technological features in 
the research and development environment. This paper assumes that 
advanced technology is available and focuses instead on the work environ- 
ment and organizational processes which facilitate or inhibit the attain- 
ment of research and developmerc objectives. 


After many decades of research, it is generally conceded that 
there is no one best way to organize. It is also agreed that not all 
ways of organizing are equally effective. If an organization is to 
marshall its resources effectively and efficiently to achieve some 
desired objective, the way in which the firm is designed should be 
compatible with the goals which are to be achieved and with the situa- 
- tional factors which provide the context for organizational activity 

(1, A, 6). Thus, organizational goals and contextual factors are two 
Important contingencies which influence what types of organization 
structures and processes are likely to be most effective in a given 

Organizational Goals 

Organizational goals serve a number of different purposes. Goals 
establish a direction for activities or describe a future state which 
the firm is attempting to realize. Goals help to legitimize the exis- 
» tence of a firm or unit, and goals provide a standard for evaluation. 

^_ Research and development units, like most other parts of an organization 

4 have many types of goals. However, because of the nature of the reseaich 

* and development unit's contribution to the activities of the organiza- 

tion as a whole, two types of objectives are of particular importance: 
i (1) providing for high quality research applications, and (2) product 

design efficiency and effectiveness. 

'" 122 




In this research study high quality application was measured by 
items such as improved engineering standards, improved service standards, 
improved manufacturing standards and decreased cost of design changes. 
These items appear crucial to developing a product which is distin- 
guished in the marketplace by exceptional quality and by responsiveness 
to customer needs. Such factors permit a firm to adopt a competitive 
strategy of product differentiation based on quality. These factors 
also affect how well the research and development unit is integrated 
with o^her functional units in the organization. Application quality is 
one indicator of the innovative capacity of the RfcD unit. 

The second type of goal, design efficiency and effectiveness, 
was measured in this study by items relating to increased efficiency in 
document production (this was one ot the primary outputs of the R&D de- 
partment), increased efficiency in design, increased effectiveness in 
design, and increased efficiency in developing product definition docu- 
ments. These items are linked with costs. The more efficiently a re- 
search and development unit is able to design and document various pro- 
ducts the fewer resources it will require to achieve a given magnitude 
of performance, or the greater its performance will be with a given level 
of resources. These factors permit a firm to adopt a cost leadership 

These two goals, (application quality and efficient, effective 
design) comprise one type of contingency considered in this study of 
designing structures and processes for a research and development opera- 

Contextual Factors 

In a research and development environment, one of the most impor- 
tant contextual factors is the extent to which the task to be done is 
understood, familiar, routine, and otherwise analyzable (2, 7, 8). 
Issues which indicate a high degree of analyzability or familiarity in- 
clude: job monotony, lack of basic Interest in the work, the belief that 
the longer an employee holds a job the more boring it becomes, a job 
situation where change is minimal, and the employee has more than ade- 
quate training and skills. Issues which suggest a low degree of analyz- 
abllity/familiarity include the feeling of challenge a job provides to 
what the employee thinks he or she can do, belief that the Job may be 
frustrating but it is never dull, and indications that something new 
happen^ on the job every day. 

On an absolute scale, a research and development environment is 
considered largely nonroutine, having many exceptions, surpirses, and 
situations which are difficult to analyze. Studies suggest that over 
ninety percent of the work in an R&D environment involves nonroutine 
technology and activities. Yet, on a relative scale, some of the tasks 
are clearly more easily understood and performed than others. Therefore, 
this contingency remains important despite the overall "uncertain" 
nature of research and development activities. 

A measure of the extent to which the work environment and subse- 
quent tasks are seen to be familiar and analyzable is the second contin- 
gency considered in this study. 



te>-.*,4. ..1 ^M»- 



The study Is based on an empirical and case-based research 
effort undertaken at a large-scale research and development division of 
a major mldwestern company. Managers, support technicians, engineers, 
and administrative employees in the research and development division 
were given a version of the Michigan Organization Assessment Question- 
naire (MOAQ) which had been modified to fit the conditions of the organ- 
ization. The sample size of 274 represents slightly over 50 percent of 
the relevant eTiployees. Organization culture dictated that participa- 
tion in the study be voluntary. As a result, the sample has dispro- 
portionately high representation from nonunion employees such as managers, 
technicians and engineers. However, data analysis shows no significant 
difference in responses between union and nonunion employees. 

In addition to the questionnaire a content analysis of formal 
organizational documents provided diagnostic infoirmation regarding goals, 
structure, organization performance and organization culture. Fifteen 
interviews with key decision-makers and multiple observations of the 
unit in operation over a period of a year and one half provide the con- 
text for the empirical analysis. 

The research and development division being studies has changed 
from a top-down, functional structure to a workgroup centered structure. 
Twentythree workgroups are identified within the division. Each workgroup 
has a unique set of goals, tasks, evaluation criteria, time-frames for 
deliverables, and relationships with other parts of the division and 
other parts of the company. Each workgroup contains a mixture of skills 
and hierarchical levels. This structure permits comparison among units 
having different work environments and workflow processes and facing dif- 
ferent task contingencies. 

Based on composite responses to items related to familiarity and 
analyzability, the workgroups were split into two categories: those 
which faced conditions of comparatively high familiarity and analyzabil- 
ity and those which faced conditions of comparatively low familiarity 
and analyzability. Eleven workgroups were classified as operating under 
conditions of lower famlliarity/analyzabllity. Twelve workgroups were 
classified as operating under conditions of higher famlliarity/analyzabll- 
ity. As mentioned previously, the scores reflect relative rather than 
absolute scales of these items. 

Every workgroup was to some extent responsible for achieving 
goals related to both quality applications and efficient, effective de- 
signs. Employee's aggregate perceptions of the extent to which goals are 
being achieved were the performance measures used in the study. A two- 
by-two correlational analysis enabled investigation of those structural 
and organizational process characteristics which facilitated or inhibited 
goal achievement under each of the two contextual conditions. 


'i Five categories of organization structure and process variables 

^ were investigated: (1) supervisor characteristics, (2) workgroup char- 

I acterlstics, (3) employee attitudes and feelings, (4) job/task character- 

istics, and (5) information processing emphasis. Each of these factors 


^ ' '^ ■(* 



riCUBt 4 Corr«)»te» of Emplogee Altitudes With Outlity tnd Efficiency CotlJ 
Under Conditions of High tnd Lov Ftmiliaritu/Analgzebilitg 















Or9 wivoWtmfflt (5^*) 


Rttponsitiimy (36*) 

Ovsirt to Q\*ngf Jobk (- 67* 

) Control (76«) 
Tgrnov»r mttnt (- 75« • •) 





positive;! ACTORS 

D*frt to ch4n9* )ob% ( 59*) 

Tgrnovfr wtoot (.60*) 


Org mvolv»m»nt(-6I») 
Rrsf onstbility (- 59) 
Control (-67««) 

♦p"05, ••p-01, •••r"005 


Conditions of high familiarity and/or analyzability suggest a 
different attitude pattern. It may be that the greatest design effici- 
ency is achieved by moderately discontented employees. Design indus- 
tries are frequently characterized by employee mobility. Further, it 
is recognized that movement is dependent on recent performance; an 
employee is only considered as good as his or her last design in many 
cases. Perhaps this interest in change and knowledge of the perform- 
ance prerequisites foster desirable engineering and design activities 
or perhaps such employees are more willing to take more unconventional 
approaches to design. High performance with regard to quality applica- 
tions appears to be fostered by a more contented attitude. Job satis- 
faction, feelings of responsibility and involvement, a more centralized, 
directive structure, and an interest in remaining in the current posi- 
tion contribute to high quality applications under familiar/anfilyzable 

Three of the four contingent conditions show a positive response 
to task characteristics generclly associated with "enriched" jobs (see 
Figure 5 ) . Only an effort to achieve efficient designs under high 
analyzability seems to be positively influenced by a more focused ana 
more loosely coupled job-task characterization. Variety and feedback 
seem to be the most Important factors overall. Quality appllca*:xcns 
appear to be aided by an interconnectedness with other units in the 
organization. This seems to complement the feelings of organization in- 
volvelment which make a similar contribution. The positive effects of 
interdependent may also indicate greater knowledge of the interests and 
fundtions or diverse operations within the firms. This knowledge in- 
creases the implementation feasibility of many research and development 




ill, .^' 


a'^. t^^^l^ 

analyzable (see Figure 3). Again the direction of influence is re- 
versed for the two types of goals. Fragmentation and heterogeneity 

riGUPf 3 Correlilej of Group Chtroctensttcs With Qoslity tnd tfficieneg Goals 
Under Conditions of High and Lov Fomiliantv/Antl'^iBbili'ig 










New Ifiitni 


Fr«9m»nU1iofi ( 76 • * «) 
H»t»rofl»n»Ug (76*»«) 

Dp»n procMsts (- 72***) 





0p«n proc'.iM i 83***) 

Fr*flm»n(«tion (- 64 « ) 
H«.«»ro9»n»it« (- 72*»*> 

p«05, **p-CI, 

• p-005 

N» 23 

have icrong positive effects on quality goals and strong negative ef- 
fects for design efficiency. Open communication among group members has 
a negative effect on quality applications yet a positive effect on design 
efficiency. This pattern suggests that work segmentation and iome degree 
of specialization may be appropriate for achieving quality applications, 
but that shared values and group cohesiveness are importa^c conditions 
for achieving efficient designs. 

An interesting pattern is evident when employee attitudes are cor- 
related with strong performance under the fcut contingent situations con- 
sidered in this study (see Figure 4 ). If employees feel challenged by 
their jobs and if they have no desire to change jobs, quality applications 
are achieved under conditions of low famillarity/analyzability. An over- 
all feeling of commitment (to the organization and to the job) has the 
strongest influence on design achievemerts under conditions of low cer- 
tainty. This suggests that in both cases internal motivation has a 
strong effect on performance. Further, it appears that application goals 
require more of a job focus, while design goals respond to a more general 
organization orientation. 




. - '-i^i.-.V; 

A supervisor's behavior end managerial style appear to have the 
strongest positive effect when promoting quality applications under 
conditions of high familiarity and analyrablllty (see Figure 2 ). Under 
this set of contingencies, supervisor's who actively encourage employee 
participation in decision-making, who facilitate subordinate interactions, 
goal setting and problem solving, who are aware of work progress and ac- 
tivities and who treat subordinates as respected individuals positively 
contribute to achieving quality applications, in contrast, these same, 
behaviors have a strong negative effect if analyzability remains high 
but the goal is to develop efficient designs. It appears that under 
this latter contingency set, artlve, facilitatlve supervisors tend to in- 
hibit performance. If familiarity/aualyatility is low, however, some 
of the active supervisor characteristics (suc'a ?b facilitating decentral- 
ized control) appear to have a positive effect of efficient design per- 
formance. Supervisor behavior did cot appear to have any influence on 
the development of quality applications when familiarity /analyzability 
is low. 

riCURt 2 CorreUtti of Vtrtous Supervisor ChtrKteristics With Qutlitv tnd 
Crrtclcncv CmIi Under Conditions of Htgh tnfi Low rtmllttrtty/ 











Nen* ETid*n( 

Control •( *»rfc ( eO»»») 
F«ci1tt«tWt roUtions (SZ*) 

IA55K fAi»t 





F»e1lt1i<W» rr i*«onj (.33«) 



Gw'. »»««in9(.76»«») 

Contrtl of vork (-.73»««) 


F«citi<«tiv* roUtwns (-.U*) 
Pro»Win lolvnf (- eo« • •) 

• p".03, ••»• 01,»«*p" 003 

N« 33 

These findings support the argument that the most effective 
supervisors may be those who are both versatile ari'^ somewhat inconsis- 
tent, effectively catching their behavior and direct involvement in thi 
workflow to each situation. These findings suggest 'har -raining super- 
visors to expand their repertoire of skills might be particularly useful 
when the task environment is fairly familiar and analyzablc, but that 
this type of investment would not have a strong effect en performance 
when the job Is performed under cuaditions of extremely low analyzability. 

Similarly, the characteristics of the immediate v;ork group have 
the strongest effect when chc situation is relatively familiar and 


■■ - 'ilf'r'^: : . 

,«fi4«ak' ■i^hK- 


was considered tor (;aality application and for efficient, effective de- 
sign goals under conditions or high or low familiarity/analyzablllty. 


Results of this study suggest an Interesting pattern of rela- 
tionships. Under conditions of high familiarity/analyzability, quality 
application goals appear to be fostered by an actively supportive and 
nuturlng work environment. In contrast, efficient design under condi- 
tions of high familiarity/analyzability, appears to thrive under a more 
lassez faire work environment. Such a lasssz faire approach may aid 
performance by removing bureaucratic impediments to performance. 

Looking at conditions of low familiarity and analyzability, "f- 
ficient, effective design appears to be facilitated by an actively- 
managed work environment, while quality applications appear to rely on 
inspiration and motivation. Figure 1 depicts the general work environ- 
ment characteristics and organization processes which facilitate and 
inhibit quality applications and efficient designs. 

f^lGUPf I Patttrn* of Org^niMtion Structure and Proc»»»«i Which Aid 

Cod Attiinment Under Conditions of High i nd Lov Ftmilitntu/ 












4*(ir« t« W«v* 



CTf «tlv* dtseonttflt 


suprrvitor in<fr«ction 

d»(irr to Way* 

cempWx ^vcKWn ttri^tur* 



>ubor4in«t» mWultan 


IA55K fAtRf 


opvn dncutston 

d*srr for mobility 




f ICURt 5 Corr«l(t*i of Job-TMkCh«rtclcri«(ic* With Quality and [fficiencv 
CmU Under Condttloni of High am) low ramiliirity/Analvzabililv 












V«rt»tM (*»••) 

V*r<»«y ( S2*) 


F»fdb»ek ( 51 •) 



mtvrn*) int»r4*p ( 64 «) 

IA5?K Jh\y\ 





V«rittv ( 62«) 



FM««ck (.58*) 


KnewrfsuHf (-97*) 

>P"03, ••p«OI, *«»p-.005 


With respect to Intormation processing, achieving efficient de- 
signs appears linked with insuring that two undesirable conditions do not 
occur. First, the manner iti whicn information processing takes place 
should not be dictated. This suggests that under conditions of high ana- 
lyzability, information processing activities should flow from the speci- 
fic tasks at hand rather than from some predetermined approach to informa- 
tion analysis. Under conditions of low familiarity, the greatest danger 
seems to come from premature analysis. Correspondingly, a second condi- 
tion to be avoided is having routine information ind information processing 
use such a largfi proportion of time or resources that none is left for non- 
routine, explorative, inventive approaches. Quality applications appear 
facilitated by insuring that adequate information is shared and available 
and that prior decisions and solutions are recorded. These results are 
presented in Figure 6 . The composite pattern suggests that efficient de- 
signs are most likely to be inhibited by information overload, while 
quality applications are most vulnerable to omissions in information. 


The correlational patterns which emerged from this study suggest 
two important considerations for the structur*' and design of research 
and development units. First, it is clear that a decision needs to be 
made whether to separate or to integrate the two primary types of goals 
most often present in these units. If the choice is made to separate 
theje activities, then diverse organization structures must co-exist in 
the same unit, and must frequently interact to achieve organizational 
goals. Separation will likely increase the need for information sharing 
and for conflict management, since developing quality applications and 



•• • •'#»"^- -. -' 

^i .v**-'5''^ 


1 . 

f ICU>it 6 Corrtlitu oT Intorfrntlen ProctJjiofl Cmph«ij With Qutlitij ind 
Cfricitntv CmIi Undtr CondUiont oi' Mi9h tni lov Ftmiluritg/ 











R(u<int »tfefKch(54*) 
Ne«rout<o* Mf* tMc (e6***: 
DocurrwnUlion ( H*) 




Rnutw* «nfo»»<e(- 7I»«») 
rionrEKitm* mfo txc (- 6« •) 

•p-05, «»p« 01, **«p-003 


efficient designs cannot and should not be self-contaiaed activities. 
If the choice is made to integrate two types of goals within a set of job 
or workgroup activities the stress of reconciling diverse objectives be- 
comes an individual rather than an organizational problem. Thus if inte- 
gration is the choice, supervisors, workgroups and individuals require 
training and experience in stress management. Further, they need to 
devslop many different sets of skills and operating styles as well as 
che ability to choose an approach to fit a given situation. 

A second issue also emerged. Researchers and practitioners ap- 
pear tc know much more about how to structure ana manage circumstances 
which have some degree of familiarity and analyzability. Despite the 
range restriction inherent in looking solely at an R^C environment, strong 
differences were found between lower and higher degrees of familiarity and 
analyzability. An important issue to be resolved, therefore, Is whether 
performance under highly uncertain and unfamiliar circumstances is almost 
exclusively a result of individual talents and capabilities or whether 
traditional measures and indices of organizational structure and process 
are just not the appropriate factors to consider. If the former 
explanation is true, selection rather than training or organization design 
oust be the dominant human resources issue for many research and develop- 
ment operations. If, however, the latter explanation is true, there is a 
need for designing creative and unconventional organization structures 
and measures to accomodate the unique environment of the research and 
development unit. 



♦i ^.i:ll'^v?^ -'.^ ■ (^ 



(1] Burns. T. and G.M. Stalker. The Kanaeement of Innovation , 
Tavistock. London. 1961. 

[2] Galbralth. Jay. Designing Complex Organizations . Add! son-Wesley. 
Reading. Hi 1973. 

(3J Gerstenfcild. A., Effective Management of Research and Development . 
Addison -Wesley, Reading MA 19 70. 

[4] Hage, Jerald and Michael Aiken, "Routine Technology, Social 
Structure and Organizational Goals," Adni.nistrutive Science 
Q uarterly . 14, 1969, pp. 366-376. 

[5] Kelly, P. and M. Kransberg, Technological Innovation; A Critical 
Review of Current Knowledge , Advanced Technology and Science 
Studies Group, Georgia Instit< te of Technology. 1975. 

[6] Lawrence, Paul R. and Jay W. Lorsch, Orgai.izations a n d Environment , 
Graduate School of Business Administration, Harvar<^, Boston, MA 

[7] Perrow, Charles, "A Framework for the Comparative Analysis of 

Organizations," American Sociological Review , 32, 1967. pp. 194-208. 

[8] Tushman, Michael L., "Managing Comis'mlcation Networks xn R&D 

Laboratories, Sloan Management Review , 20, Winter 1979, pp. 37-49. 


:'i 131 

.- • ^if^' 

i^\»*4/^'^;.V;vv.v-.. .. 

■y I 



Richard L. Davis, TRW 
Donald A. Molgaard, TRW 


The Gamma Ray Observatory (GRO) Program has been proclaimed to 
be the sbowc-'se productivity program for NASA and TRW. Among the mul- 
tiple disciplines of a large-scale program, there is opportunity and 
need for improved efficiency, effectiveness, and reduction in the cost 
of doing business. This paper describes the efforts and tools that 
will or have been implemented to achieve this end. 

Since the GRO Program is mainly an engineering program with the 
build of one satellite, the primary emphasis is placed oa improving the 
efficiency and quality of management and engineering. 

Top management of TRW, NASA/Headquarters and GSFC are totally 
committed and firmly endorse productivity for the GRO Program as shown 
by their willingness to implement potential cost saving tools and re- 
placing older operating procedures with new ones that take advantage 
of today's technologies for more efficient performance. 

During the initial part of the Gamma Ray Observatory (GRO) 
Phase D contract, the project designed and developed a high-fidelity 
full-scale model (FSM) of the GRO. In addition to its use as a de- 
sign, fit check, and personnel handling and training aide, the prin- 
ciple design objeccive for the FSM was to evaluate (prior to the Cri- 
tical Design Review) the current design in terms of its suitability 
for performing both planned and contingency extravehicular activity 
(EVA) operations in the deployment, repair/refueling, and retrieval 
missions of the GRO. Typically, spacecraft development programs have 
not performed this type of EVA design evaluation tests at such an 
early stage of program development. Normally the EVA evaluation ex- 
ercises are performed as part of the astronaut crew training opera- 
tions 3 to 9 months prior to launch. A program could incur signifi- 
cant cost/schedule impact from design deficiencies identified this 
late in the program. By addressing the EVA design compatibility 
early in the program, the GRO project was able to accommodate pro- 
posed changes with minimal cost and no schedule impact. 


j;ff*w'^i«f««-»>- • 


TRW is building the GRO platform to carry four large instru- 
ments to conduct a full sky survey and study selected objects of in- 
terest in the gamma ray region of the electromagnetic spectra. The 
Observatory is 25 feet long, 15 feet wide and 12 feet high weighing 
over 34,000 pounds. The solar arrays span 70 feet. Figure 1 shows 
the GRO configuration. 









Figure 1. Gamma Ray Observatory 

The GRO will be launched by the Space Shuttle and placed in a 
low Earth's orbit for a minimum of 2 years when it will either be re- 
trieved or refueled by the Shuttle. 


\ \^ 

--:. I 


Within TRW, productivity is an integral part of the organization 
and operation. The GRO Program has a Productivity Manager who reports 
directly to the Program Manager. He is also a member of the Federal Sys- 
tems Division Productivity Council which reports to the Division Vice 
President. This Council reports to the Space and Technology Group Pro- 
ductivity Council. Through this network the productivity actions of all 
the groups are coordinated to ensure that the maximum benefit is passed 
on to all organizations. 

Initial Actions 

At the beginning of the program, a one day brainstorm session 
was held with program management to develop productivity ideas. From 
this session macro goals were established. These were presented to the 
NASA Productivity Steering Committee on 13 July 1983 along with a com- 
mitment for implementation within one year. 

Simplified Performance Measurement System 

In order to determine the status of a Work Breakdown Structure 
(WBS) element, work package and /or task, the manager must wade through 
a large computer printout which is not only time consuming but often 
does not get done properly due to the magnitude of the task, A simpli- 
fied automated :nethod was required to ensure quick and accurate assess- 
ment of the program status. 

The Air Force developed such a method for the MX program which 
was called Red Flag. The PMS data is inputted into PCs and the output 
displays are then available for any level of the WBS. These outputs 
are color coded as a function of variance levels for easy identifica- 
tion of problem areas. The trend from the corrective action can be 
taken immediately. Displays are available for cost and schedule var- 
iances, variance at completion, manpower plots, performance factors, 

Automated Critical Path Schedule Network 

Several years ago Pert was the preferred method of tracking a 
program schedule status. However due to the many hours required to 
manually draw the network(s) and status the critical path. Pert was 
dropped in favor of Gantr. charts which could be easily drawn and even 
automated. Put Gantt charts did not provide all the necessary infor- 
mation. Today several automated Pert programs are available. TRW is 
utilizing the Project 2 Scheduling System. This is a totally automa- 
ted network schedule system coded to provide traceability to the low- 
est level planning milestone in the PMS, It is to be updated monthly. 
It features critical path networks, logical related bar charts, sub- 
nets with multiple work calendars, interactive graphics and has a 
"what if" analysis capability. 



Improved Communications 

TRW must interface with GSFC, NASA Headquarters, four major in- 
strument teams, McDonnell Douglas (MTG), Fdirchild (CADH) and four major 
subcontiactors all of which are located away from TRW. TRW and GSFC are 
located 2500 miles from each other. An Improved method of communication 
was required to increase efficiency, effectiveness and motivation while 
decreasing the cost of doing business. 

Two method have been incorporated to achieve this goal, a com- 
puterized network system and video teleconferencing. 

Computerized Network System 

TRW and GSFC jointly analyzed and selected the most cos effec- 
tive personal computer system for improving not only communications but 
also to improve office efficiency. The DEC Rainbow 100 was selected as 
offering the most capability, highest speed and best operation at the 
lowest cost. 21 work stations have been Installed at TRW and 11 at GSFC. 
These stations have either letter-quality or matrix printers and high 
speed modems for inter-computer communications. 

The initial software packages procured were word processing 
(Wordstar), electronic spread sheets (Multiplan and Lotus 1-2-3), data 
base management (Dbase II) and communications (Mite). One cost effec- 
tive use was the development of computer generated graphic presentations 
which is used for all design and management reviews. 

Training classes have been held on the use of the computer and 
each of the software packages. These classes will continue to train 
new personnel and teach the operation of new software packages as they 
are acquired. 

Video Teleconferencing 

A normal meeting time span is approximately 2 hours involving 5 
to 10 people. When travel is required to attend a meeting 2 to 3 days 
are spent away from the office. Full-frame video teleconferencing for 
technical and program meetings can reduce the travel expenses and time 
away from the office thus improving program efficiency. 

Several test meetings have been conducted using rented facili- 
ties in Los Angeles and Washington D.C. The success of these meetings 
convinced both TRW and NASA to install video teleconferencing facili- 
ties at both locations. 

Maxi mize Technology Transfer and Lessons Learned 

GRO is maximizing the technology transfer and lessons learned 
not only among TRW programs but also amoung other NASA programs. TRW 
is a matrix organization where except for some of the project manage- 
ment the majority of people are supplied by functional groups. These 
functional groups support not only NASA programs but also defense and 
commercial programs. As a result the technical achievements on these 
programs can and will be used during the GRO development, manufacturing, 
assembly and test phases. 




The NASA personnel will supply technical guidance learned from 
other programs as well as supplying safety, reliability and design a- 
lerts. The NASA design review team is composed of members who have ex- 
perience on many programs to ensure that past mistakes are not repeated. 

Improved Procurement Cycle 

A reduction in paperwork and rapid responses turnaround time 
will greatly enhance productivity between GSFC and TRW and between TRW 
and our subcontractors. Three procedures have been or will be imple- 
mented to achieve the NASA goal. 

The NASA 533 have been automated reports using the company PMS 
and converting to the NASA format. A productivity incentive clause has 
been implemented which will allow TRW to obtaina higher fee based on 
ideas for measurable cost savings. For every major cost saving gener- 
ated and approved by NASA, TRW will retain 20 percent. If the program 
is overrun at launch, 50 percent must be returned. If there is an on- 
orbit failure that was caused by TRW, 50 percent must be returned. A 
portion of this will be distributed to the GRO employees. 

The TRW GRO Program has prepared a subcontractor communications 
plan which flows down the program requirements, including lessons 
learned, and utilizes TRW's computerized network. Each subcontract 
contains the product assurance requirements which include previous TRW 
spacecraft programs lessons learned. GRO subcontracts is providing 
these lessons to subcontractors in request for proposals to facilitate 
their learning and reduce costs. The GRO Program Office will perform 
design reviews and materials and process audits to assist subcontractors 
in applying lessons learned. TRW will implement utilization of person- 
al computers and freeze-frame video conferencing for selected subcon- 
tractors. When fully implemented, these productivity improvements will 
reduce the time and cost of manually prepared cost and schedule reports, 
personnel travel expense, and cost of message transmission. 

TRW did hold a subcontractor productivity seminar on 12 and 13 
January 1984. 13 potential subcontractors were invited and 10 accepted 
and participated in the meeting. 43 productivity ^'^cis were generated 
and are being evaluated for implementation. 

Individual Av»ard System 

Improving productivity and enhancing personal motivation e- 
quires an individual recognition/reward program which bar. been imple- 
mented. Recognition is achieved through two ways. The first is the 
GRO Briefs, the monthly program newsletter, which recognizes achieve- 
ments of individuals who have contributed in a productive manner. The 
second is letters of recognition which are issued by GSFC, GRO Program 
Office, and/or Federal Systems Division Manager for individuals who 
have performed outstanding accomplishments. 

Individual rewards are achieved through three ways. The first 
is that any individuals contributing to a program cost savings of more 



p.^^kCV,''* ' \\ 


Chan $500,000 will receive a GRO model at an all-hands meeting. The 
second Is each award fee period (every 4 months). Individuals, SFM or 
below, who the Program/GSFC feel contributed the most to the program 
will receive a GRO model at an all-hands meeting. The third is dur- 
ing the year, TRW will continue its policy of rewarding a monltary 
bonus to individuals whose accomplishments warrent special attention. 

Quality Circle 

A Quality Circle has been formed within the GRO Project. It 
is headed up by the SFM for Thermal Design and meets once a week. 
Typically they lock at the day-to-day operations to see if and where 
improvements can be made. Several positive ideas have been developed 
to date. Additional circles are envisioned as the labor mix changes 
in later program phases. 


In close coordination with NASA/ Johnson Space Center (JSC) 
personnel, a series of GRO EVA design evaluation tests were sched- 
uled and performed in the MASA./JSC Weightless Environment Training 
Facility (WETF) using the GRO FSM during the period from 13 February 
to 5 April 1985. Secondary objectives of these GRO FSM WETF test 
operations were to: 

o Identify the need, if any, for special hardware 
bumper/shock protection on critical GRO compon- 
ents within the planned EVA work or translation 

o Validate the compatibility of the integral GRO 
berthing adapter with the GSFC FSS-developed A 
prime cradle scheduled for use on any GRO repair 
or refueling mission. 

o Verify that all identified planned and contin- 
gency EVA opei-ations can be performed using ex- 
isting standard EVA tools and equipment. 

o Familiarize the JSC flight crew personnel with 
GRO EVA operations to allow conceptual formu- 
lation of EVA scenarios. These scenarios could 
be reviewed and fine-tuned prior to the formal 
GRO EVA crew training operations scheduled in 
1987, approximately 7 months before launch. 

FSM Design and F abrication 

The GRO FSM was designed to be compatible with the water immer- 
sion environment of the JSC WETF. All materials and coatings used have 
proven tolerance to prolonged exposure to the chemically treated water 


•\ •7«^' 

1 \ ^v ^1^ Sv > 

in the WETF. Nj wood or wood products vere used in the FSM. Primary 
structural elements were made of aluminum, with the secondary struc- 
tures, expetiments, and components made of Lexan. The FSM was fabri- 
cated by i> vendor outside TRW from engineering drawings and pre-release 
flight drawings available at the time of PDR. The FSM was returned tc 
the vendor for upgrade to the CDR configuration prior to the JSC WETF 
tests. The FSM was shipped partially disassembled from TRW in Los 
An^pies to the JSC WETF by commercial air-ride double-drop trailer op- 
.iiating with a wide load permit. 

Crew-Supported Testing 

The GRO FSM was reassembled at JSC and transported to the WETF 
on 25 February 1985. An initial series of pretest EVA evaluation ex- 
ercises were performed b> support personnel in scuba gear. The GRO 
FSM was subjected to five separate astronaut crew runs. These tests 
and the participants are identified in Table-1. The configuration of 
the GRO in the WETF for these tests is shown in Figure 2. 

GRO/FSM EVA WETF Operatio ns 

As shown in Table 1, the individual tests were performed to 
support specific objectives for t'.ie deployment /retrieval and the 
repair /refueling missions. Because of the large size of the GRO and 
the relatively shallow depth of the WETF, a portion of the GRO pro- 
truded above the surface of the water. In most Instances, however, 
this did not compromise or invalidate the test. 

Deployment /Retrieval Mission EVA Simulation 

The solar array and high-gain antenna appendage deployment 
mechanisms (Figure 2) on the GRO are designed for automatic motor- 
driven deployment initiated by ground command from the GRO/POCC after 
the GRO is out of the cargo bay but prior to release by the RMS. 
Should any of the latch actuators or drive mechanisms fail to oper- 
ate normally, each appendage mechanism is equipped with a back-up, 
EVA-operated mechanical override 'capability that allows the suited 
astronaut in EVA to deploy the appendate using the standard EVA rat- 
chet wrench in the orbiter tool inventory. In the unlikely event 
that the astronaut is unable to complete the EVA override appendage 
deployment operation, the design will allow the astronaut to restow 
and relatch the appendage prior to GRO return to the cargo bay for 
troubleshooting or return to earth. If the appendage mechanism has 
failed in a partially deployed condition and cannot be deployed or 
restowed, the appendage can be Jettisoned by EVA action. 

These EVA appendage operations were evaluated in two series 
of tests with different astronaut subjects. Based on comments re- 
ceived during the test and at the post-test critique, a recommenda- 
tion was made to incorporate additional handrails and foot-restraint 
sockets to improve EVA accessibility at the various work stations. 
In addition, a recommendation was received to reclock the three 
solar-array jettison bolts for improved EVA tool accessibility. Thtise 




>l4XjLBIk. \. ■ 









































































cd o 

Vi 1-t 

(d w 

iH 4J 
O 01 









0) >% 

i-l O >«-l 

3 rH O 

O 01 4-1 


3 < O. 

Od Q Oj 

o a Ptf 






iH U 





4J <e 





«0 rH 






r-t O 






99 09 









« "O 





Vl 01 





W >v 




<— 1 






3 '-' 






ec o. 





O 0) 






3 >> 






<U 4J 




01 iH 
VI rH 









•o > 

4J Xi 





1-1 i-( 




cd CO 

.O 0) 




_, *• 


Vl C0 





SB 4j 


O 0) 





a 3 
< o 
u VI 



1 o 
























1-t 0) 


1-1 0) 



•-t 01 



F-l 0) 




r-l I-) 



r-l rJ 










M • 



W • 






• • 


• • 



O X, 



o X 











ix! • 


ifi • 






• 4J 



• tJ 




Vl a 



M P. 




Q o 



O O 





























:§ S 







.— I 

— ^ 



















*i .^'.a'^^v^ 










reconmendations were Incorporated InCo the GRO FSM and were evaluated as 
part of the subsequent tests on March 18 and 19, 1985. The additions 
and modifications were; approved by the test crew personnel. These mod- 
ifications have been sacorporated into the flight design. 

Repair/Refueling Mission EVA Simulations 

The GRO design incorporates three electronic subsystem modules 
that are capable of being removed and replaced on orbit by EVA. These 
Multi-mission Spacecraft (MMS) standard modules developed by NASA/GSFC 
are flight qualified and currently in use on the Solar Max and Landsat 
missions. The two GRO MMS power modules and the MMS communication and 
data handling module is identical in physical form and fit to the MMS 
module that was successfully replaced cci the recent Solar Max Repair 

The EVA simulations established EVA translation routes for 
handling of the largp. modules with a minimum of transfer and hand-off 
operations. Recommendations were received and Incorporated for adding 
additional handrails on GRO. 

The GRO is the first U.S. spacecraft to Incorporate an on-orbit 
refueling capability. The on-orbit refueling coupler is being devel- 
oped for JSC by Falrchild Controls. A mock-up of this coupling was 
provided for EVA evaluation with the GRO FSM. Several recommendations 
on handle placement and the ramping of the latch mechanism were made 
during the EVA simulation tests. 


The EVA design evaluation tests performed by the GRO project 
using the FSM significantly reduced the possibility for costly, time- 
consuming modifications that otherwise might not have been identified 
unlcl the astronaut EVA crew training operation at 3 to 9 months before 
launch. Required design modifications observed during these GRO FSM 
WETF activities have been incorporated into the flight design with aii- 
nlmal cost impact and no schedule impact. The JSC astronaut crew per- 
sonnel and their support planning organizations have become familiar 
with GRO at least 2 years before the final E' A astronaut training op- 
erations are scheduled. With the knowledge and hands-on experience 
gained by all participants irt this initial operation, the final El'A 
training operations should be much easier and minimize significantly 
the possibility of costly real-time rework and retest. 


... r-#5»«^. 




Wayne D. Bodenatelner 

Assistant Professor 

University of Texas at Arlington 

Former Deputy Chief Naval Material for Acquisition 

Edvln A. Gerloff 

Associate Professor 

University of Texas at Arlington 


This paper describes certain structural changes in the M «1 
Material Coimand vhlch resulted fron a comparison of Its operations 
to those of svlectec large-scale private sector coapanles. Central 
to the change was a reduction In the: number of formal reports from 
systems commands to headquarters, and the provision of Program Manage- 
ment Assistance Teams (At the request of the pro|(ram manager) to help 
resolve project probIea(«. It Is believed that these changes Improved 
conmunlcatlon and Information-processing, reduced program manager 
stress, and resulted In Juproved productivity. 





The Naval Material Establishment Is responsible for the develop- 
ment, acquisition, and support of all the complex weapons systems needed 
by the Navy to meet its world-wide commitments. The procurement of 
Naval weapons systems often involves substantial advances In technology 
because the equipment is required to operate in extremely diverse and 
hostile environments. Advances In technology, in turn, mean that a 
given program manager (PM) must contend with high levels of uncertainty 
in managing the assigned project. Ideally, the FM seeks a technically 
sound weapon, delivered on time, and within budget. Practically speak- 
ing, however, the PM finds these objectives to be somewhat in conflict 
with one another. When coupled with the Inherent uncer' ^inty of R&D, 
the simultaneous accomplishment of technical performanc , schedule, and 
cost objectives are powerful sources of stress for the PK. Further, 
the substantial uncertainties of P^&D create an extraordinary Hemand for 
intra- and inter-organizatlonal communication If the project s to he 

Critical Issues for the Naval Material Comr.and (NMC) and the PM 
In managing such high technology are concerned with' how effectively their 
organizations are i\ble to communicate the information needed to deal with 
uncertainty, and also how well individual PM's are able to cope with the 
stresses of their jobs. This paper reports the resulr.s of a case study at 
NMC. Specific attention is given to certain organizational changes made 
by executives at NMC to help PM's Improve communication and cope with the 
stress of thtir projects. It is believed that these changes Improved 
overall program productivity. The case presented here was part of a 
broader investigation which involved a comparative analysis between NMC 
and selected major private sector corporations. The specific purpose of 
the investigation was tc improve the efficiency of NMC and its headquarters 
staff (Gerloff, 1985), 


Tne NMC was originally organized in 1966 to centralize Naval 
procur<;ment and R&D. At its Inception, NMC was organized in a divisional 
or systems command format which included Naval air, sea, electronics, and 
supply divisions (commands). The new structure was comparable to many 
large commercial organizations and resulted in an improvement in the 
Navy's diverse acquisltior effortrf. However, by 1980 there were some 
indications that NMC's structure was no longer quite as efficient as it 
had been initially. 



The headquarters staff was now very large (In excess of 1000 
people). Executives believed NMC was becoming too centralized, too 
concerned with Its own Issues, and too slow to respond. To some extent, 
top management was flr""lng Itself bogged down in problems that could 
better be handled at the system connnand level. Further, there were some 
Indications that this was interfering with the efficiency of the systems 
commands. System command personnel were unable to give their full 
attention to developing and acquiring the technology needed by the Navy. 
Some of their tim3 was consumed in preparing reports for NMC headquar- 
ters. Though no major difficulties had yet occurred, executives at NMC 
wanted to tighten their system and head off any problems that might be 
developing. To accomplish thir^ purpose, a comparative analysis was 
undertaken to see how the NMC headquarters staff operation compared to 
that used by large private sector firms. Over a 6-month period, NMC 
headquarters officials interviewed executives and observed the 
headquarters opera ^ons of several major U.S. corporations. 


On their return to Washington, executives at NMC used the 
information gathered in the field study to analyze the headquarters 
operation at NMC. They concluded that the managerial approaches of the 
ssveral private sector companies were very similar to each other and very 
different from that used at NMC in four ways (Gerloff, 1985, p. 273): 

1. All were decentralized and had small headquarters staffs. 

2. All used both long- and short-range corporate planning 
(strategies) • 

3. All paid very close attention to the management of their 
terminology base, in particular, the early development stages 
of new products. 

4. All managed their operating divisions via the careful alloca- 
tion of resources and maintained oversight and control with 

a minimum of upward information flows. Needless reports and 
excessive interference were avoided. 

Corrective Measures 

In view of these findings, several important changes were made in 
the NMC headquarters structure and operations (Gerloff, 1985, p. 27A) . 
The size of the headquarters staif was reduced (from over lOOO to about 
500 people) and the role and scope or its operations were reduced. 
However, special emphasis was given to managing the technological base. 
The volume of upward reports from the systems commands (to NMC head- 
quarters) was cut by 90 percent. Each system command was made fully 
accountable for its individual mission. An NMC Board of Directors was 
established which included the heads of the systems commands in Its 
membership. An NMC corporate plan was developed to guldi» operations 
over the long term. 


5. ♦) 


Against chis background of general change in the structure and 
operations of NMC headquarters, certain additional changes were intro- 
dure-j which were specifically beneficial to the individual program 
managers as they carried out their missions. The job of a PM is made 
difficult by the high uncertainty of R&D and the consequent need for 
problem-solving information and communication. Simultaneous pressures 
to meet changing operational requlrementSt project schedule deallnes, and 
cost limits while solving complex technical problems often mean che PM 
must also endure severe levels of stress. 

Special Program Management Assistance Teams (PMATs) were estab- 
lished at NMC which were instrumental in helping the PM access needed 
problem-relevant information while also coping with the Inherent stresses 
of program management. Each PMAT consists of a group of experienced, 
former program managers who are available at tha request of a given PM. 
The PMAT can provide consultation, added expertise, assistance with 
special problems, or program assessment depending on the desires of the 
individual PM. The PMAT reports only to the individual PM, and no 
written reports are used unless requested by the PM. 

Benefit to the Program Manager 

Executives at NMC believe the PMATs have been extremely effective 
and have Improved the efficiency of the various programs. They have 
also been well received and used by the PMs themselves. Further, the 
PMAT concept seems to be soundly based on communication and behavioral 
science research. For example, the communication literature has long 
argued the benefits of using the richer face-to-face communication 
channels when dealing with complex problems (Bodensteiner, 197'.: Woffjrd, 
Gerloff, Cummins, 1977; Gerloff, 1985). Further, the behavioral science 
literature suggests that an affiliation with others can help indivi- 
duals to cope with high vinxiety or stressful situations (Schacter, 1959). 
Thus, the PMAT presented a troubled PM an opportunity to discuss problems 
with people who (1) understood because they have experienced similar 
problems before, and (2) are technically knowledgeable and able to 
Introduce additional problem relevant information. It should be emphasized 
that such face-to-face discussions are "/aperior to more formal and less 
rich communication channels, especially where complex problems are 
Involved (Bodensteiner, 1970: Dirt and Wigenton, 1979; Wofford, et al, 
1977). A side benefit ^s that the technical experts, who had over the 
years moved from the systems commands to the headquarters staff, gained a 
new sense of accomplishment. By serving on PMATs, they were able to use 
their technical know-how in a way that ordinary staff work often did not 

Senefit to Top Management 

Beyond such valuable assistance at the program level, NMC found 
[^ that PMATs were also beneficial to higher management in its effort to 

y* manage the technological base. The management literature Itdlcates 


^ :'"-r*5:^'- 


.-A.» >•- 

that an important part of .nanaglng high technology Involves the need 
for top managers to be active in the early phases of a project (concept 
foripulation, design, and development). Decisions made in the early 
phases will have long term and high dollar impacts (Gluck and Foster, 

Executives at NMC found that they could use the expertise of 
PMATs in the e?.rly assessment of a new program, about 6 months after 
stavt-up. This would be before any request^ for assistance by a PM, 
and was not associated in any fashion with the primary function of the 
PMaT as described previously. At this early juncture in a program, it 
is cri'cical for higher management to assess whether adequate dollars, 
personnel, and other resources have been made available to the PM for 
successful program completion. The PMAT has proven invaluable to this 
early assessment. 

At the same early critical juncture, the PMAT can be used to 
assess whether operations at the program level are organized to effi- 
ciently use the resources allocated by higher management. The point of 
the assessment being to determine the likelihood that the program can 
produce the desired technology on time and within budget. Though this 
phase of the analysis can be a threat to the PM, NMC executives do not 
feel that it has interfered with their (the PMs) use of the PMAT In later 
phases of the project life cycle. The overall benefit of such an early 
assessment of resource allocation by top management and efficiency in 
resource utilization in a given program enhances the likelihood that a 
poorly conceived prograa can be scrubbed where circumstances warrant. 
Such a use of the PMAT concept opens the possibility that top managers 
and PMs alike will f-'nd It easier to cope with the stresses of scrubbing 
a troubled program before the sunk costs (in terms of both dollars and 
psychological costs) are too high. 



Wayne D. Bodenstelner Is an Assistant Professor at the University of 
l Texas at Arlington and received his Ph.D. In management and operations 

^ research from the University of Texas at Austin (1970). Prior to his 

^ retirement as a two-star admiral, he was a ranking military expert In the 

Navy's research, development and acquisition process, administering more 
than $40 billion In development and acquisition programs. 

y Edwin A. Gerlof f Is an Associate Professor at the University of Texas 

at Arlington and received his Ph.D. In management and statistics from 
the University of Texas at Austin (1971). Prior to joining the Univer- 
sity he was employed by the American Telephone and Telegraph Company. 
He has published articles In IEEE Transactions on Engineering Manage - 

' ment . Engineering Management International , and Joarnal of Applied 

Communication Research . 










[1] Bodenstelner, W.D., "Information Channel Utilisation Under Varying 
Research and Development Project Conditions: An Aspect of Inter- 
organizational Communication Channel Usage," doctoral dissertation, 
The University of Texas at Austin, 1970. 

[2] Daft, Richard L. and John C. Wiglnton, "Language and Organization," 
Academy of Management Review , Vol. 4, No. 2 (April, 1979), pp. 

[3] Gluck, F.W. and R.N. Foster, "Managing Technological Change: A Box 
of Cigars for Brad," Harvard Business Review , Vol. 53, (September- 
October, 1975) pp. 139-150. 

[4] Gerloff, E.A., Organizational Theory and Design; A Strategic 
Approach for Management , McGraw-HlTl, Inc., 1985. 

[5] Schachter, S., The Psychology of Affiliation: Experimental Studies 
of the Sources of Gregarlousness , Stanford University Press, 1959. 

[6] Wofford, J.C., E.A. Gerloff and R.C. Cummins, Organizational Communi - 
cation: The Keystone to Managerial Effectiveness , McGraw-Hill, Inc., 



. -_'*«-^. 



Mildki. V^ :. 





... f , 





G. Montoya 
P. Boldon 

McDonnell Douglas Technical Services Company 


The Space Station Program is highly complex not only in its tech- 
nological goals and requirements but also in its organizational 
structure. Eight Contractor teams supporting four NASA centers plus 
Headquarters must depend on effective exchange of information — the 
lifeblood of the program. The Technical and Management Information 
1 System (IMIS) is the means by which this exchange can take place. Value 

- of the TMIS in increasing producti"ity comes primarily from its ability 

to make the right information available to whomever needs it when it is 
' needed. This paper addresses productivity of the Aerospace professional 

and how it can be enhanced by the use of specifically recommended tech- 
niques and procedures for information management using the TMIS. 


Mercury, Gemini, Apollo, Skylab...The Space Shuttle. Next in that 
series of technology expanding endeavors is the Space Station. Where 
previous spacecraft were intended for relatively short visits to space, 
" -the Space Station is intended to provide the base for a permanent 

presence of humans in space. The key characteristic that makes the 
Station different from previous spacecraft is that, rather than a means 
for transportation, it is a laboratory and a factory in the sky not 
intended for a single, short duration mission but expected to remain in 
service for an indefinite length of time that spans well ""'nto the next 

The Station includes a variety of systems and depends on know- 
ledge from a multiplicity of disciplines. It is composed of a structure 
that supports experiments and space production equipment, pressurized 

♦ habitation and laboratory facilities, utility equipment such as power 
<- generation and communications equipment, and Station operations support 
4 facilities such as satellite servicing provisions and remote manipulator 
'I arms. Figure 1 shows one of the conceptual configuration now being 

* studied. 


•^ - '^^vr^^r '. 

\ » 








The Space Station Pro»ram 

' The Space Station Program (SSP) is currently in the early stages 

r of Phase B, Definition and Preliminary Design. The overall program 

f schedule leads to an initial operational capability (IOC) in 1992. The 

f program management structure is shown on Figure 2. The program is 

k divided into A work packages, each covering a set of end items and func- 

tional responsibilities under the management of a separate Center. Each 

• center has under contract 2 separate contractors (or contractor teams) 
performing parallel Phase B studies. 

* The technical responsibility and geographical distribution of 
prime contractors (and their subcontractors^ the Level C NASA centers, 

^ the Level B program management, and the Level A Agency Office create a 

complex network of information users and generators. This network feeds 

and thrives on the data that flows through it and the information dis- 

1^ tilled from this data. It is the proper acquisition, distribution, and 

processing of information that lead to the analysis, design, implementa- 

f tion, and successful operation of the Space Station. 






The Problem; Making Information Readily Available 

:t A key programmatic goal is the use of innovative concepts to in- 

"> crease productivity in the design, implementation, and operation of the 

I Station. The complexity of the Station system requires large amounts of 

'. data to be processed and information to be analyzed. The diversity and 

' geographic distribution of program participants demand that this 

',■ information be made available in many places, at all times, and in 

several formats at various levels of detail. Streamlining of information 

exchange in this complex environment can significantly contribute to 

achieve the desired goal of high productivity in the Space Station 


The Solution ; The Technical and Management Information System 

NASA recognized the need for expedient distribution of and access 
to accurate, current information and specified in the Phase B Statement 
of Work that all contractors should use the Technical and Management 
Information System (TMIS). All NASA centers, contractors, and major 
subcontractors are required to exchange certain types of information and 
documents by electronic means via the TMIS, always striving to do 
business in a reduced-paper manner. 

The TMIS is a distributed network of data processing nodes 
located througout all the facilities, government and private, of the 
organizations involved in the Space Station Prograu. The system includes 
the communications equipment and all the support software. NASA is in 
the process of developing the core and backbone of TMIS while each 
contractor is expected to develop its own segment of the overall net- 
•»j work. Figure 3 shows a conceptual diagram of TMIS. 


. V'feSf^- 









1 " 



2l -i 










? < 

O ^ 


? < 







I. t 






There has been a significant increase in awareness of the need 
for productivity improvements in all aspects of American industry during 
the last decade or so. This may be a response to what some economic and 
social commentators have referred to as the "Japanese Challenge". This 
awareness has led to several studies on productivity in different work 
environments. Manufacturing productivity has been of major concern to 
industries such as automotive and electronic production. The following 
studies deal with white collar or office productivity and they are 
reported here because of their relevance to Aerospace worker 
productivity, specially in an R&D environment. 

The Hughes Aircraft Study 

An extensive and continuing study of R&D productivity was under- 
taken by Hughes Aircraft Company from 1973 to 1977. Findings of the en- 
tire five year study effort were documented in a report entitled R&D 
Productivity (Reference 1]. The study concentrated on identifying fac- 
tors most likely to impact productivity in an R&D environment and then 
determine what techniques help to counteract the effect of each factor. 
Table 1 lists the 25 most significant factors from this study. 

These factors were then analyzed through surveys, interviews, 
study groups, and expert consultations to determine how each factor de- 
tracts from productivity. Results of this analysis were used to formu- 
late a series of techniques, procedures, and organizational characteris- 
tics that can bring about significant increases in productivity when 
incorporated in the work environment. For 19 of the 25 counterproductiv- 
ity factors, effectiveness of the techniques applied depend piimarily on 
improving information dissemination and facilitating information ex- 
change. This should not be surprising when considering that the produc- 
tivity of an aerospace professional depends on that individual's ability 
to obtain, ass^ni'ate, and issue information. Figure 4 illustrates the 
pivotal role of (he scientist, engineer, or manager in the flow of 
information. The obvious conclusion is that the single most effective 
element in increasing productivity is the provision of a system for 
expedient information flow. The THIS is such a system for the Space 
Station Program. 


A more recent study on the effect of information flow on produc- 
tivity of th • aerospace professional was conducted at thi. Johnson Space 
Center Missiori Operations Directorate. The purpose of this study was to 
determine the need and formulate requirements for an information network 
to be used by the Directorate as it moved from the R&D phase to the high 
flight race, operational period of the Space Shuttle program. Th re- 
sulting system was the now operational Management Information Database 
Automation System (MIDAS), 

The study, conducted in 1983 by McDonnell Douglas under ^he engi- 
neering and operations support contract for NASA, was based on the 




Table 1 
25 Factors Most Likely To Causa Serious 
Counterproductlvity Within RW) Organizations 

• 1. 

• 2. 

• 3. 

• 4. 








Ineffective planning, direction, and control 
Overinflated organiztion structures 

Insufficient management attention to productivity, and to 
the identification and elimination of counterproductive 
factors within the organization 
Poor Internal communicafon 
lnadeq(*ate technology 

insufficient or ineffective investment in independent 
research and developnvwt (IR&D) vjfforts 
Poor psychological w/ork environment 
Lack of people-orient&tioii in manage, ont - Insufficient 
attention to employee motivation 

Ineffective structuring o' assignnients 
Lack of effective performance appraisal and feedback 
Insufficient attention to low producers 
Technological obsolescence 

Ineffective reward systems which inadequately correlate 
individual productivity and compensation 
Lack of equitable paralbl managerial and technical 
pronrtotion ladders 
Lack of equity in operations 
Ireffectivj customer interface 
Ineffective engineeringf/production interface 
Ineffective subcootractor/supolier interface and conttol 
Operational ovsrcomplexity - constrictive procedures and 
red tape 

Excessive organizational politics and gamesmanship 
Excessive provincialism 
Ineffective management development 
Inadequate investnnent in, and lack of (tropv maintenance 
of, capital facilities 

Indicates where improving information excha'ige will alleviate 
the impact of these factors. 








E < 




\^S^ ■ 



techniques of the IBM Business System Planning approach. This approach 
emphasizes two key factors in the successful development of an informa- 
tion sytem: 1) Commitment and support by top management, and 2) User 
orientation and involvement in the definition of system requirements and 
architecture. To satisfy the second factor, individual interviews were, 
held with 35 managers and individuals whose job was data intensive, that 
is, people who spend the major part of their work day dear ing with facts 
and figures, often stored in computers or stacks of record keeping 
books, such as personnel staff and training activity schedulers. Table 2 
lists the questions that were asked of each individual interviewed. 

Table 2 
Survey Questions For MIDAS Study 

1. What is your area of responsibility? 

2. What are the main objectives of your job? 

3. What are the three greatest problems you have met in achieving these 
objectives within the last year? 

I 4. What has prevented your solving them? 

5. What is needed to solve them? 

6. What value (in man-hours saved, dollars saved, or programs enhanced) 
would better information have in, these aresis? 

7. In what other areas of your responsibility could greatest 
improvc-nients be realized, given the needed information support? 

8. What v'ould be the value of these improvements in man-hours saved, 
dollars saved, or programs enhanced? 

9. How would you rate your information support with respect to 
adequacy, validity, timeliness, consistency, cost, and volume? 

10. What is the most useful information you receive? 

11. How are you measured? 

12. How do you measure your subordinates? 

.3. What other kinds of measurement are you expected to make? 

14, What kind of decisions are you expected to make? 

13. What major changes are anticipated in your area in the next year? 
Three yeais? 

16. What do you expect l:o result from this study? 

17. Do you have any additional thoughts or comments? 

The most sip.aificant finding relating to productivity and informa- 
tion flow from this study was that aerospace professionals spend from 50 
t' 90 percent of their v;,uk day searching for information. This condi- 
tion leaves less than half and in some cases as little JS one tenth of 
available time to operate on that information and be "productive". 

Summary Findin g 

In attacking a management problem, it is important to identify 
the vital few factors that cause the greatest effect on the problem and 
then focus on those in reaching a solution. When the problem is tne need 
to improve productivity, the above studies point to a single factor 
whicn can produce I'he greatest success: facilitating information flow 
Throughout the organization. Therefore, any system which can make infor- 
mation available to whomever needs it, when needed, and in 5 he form 
needed will have a sif.r.if jcant effect in increasing productivity. The 




Sp?xe Station Program took a significant step towards meeting its 
productivity goals when it conunitted to the development of the THIS. 

NASA Direction on Use of TMIS 

NASA has made an agency-wide commitir.ent to use TMIS on all possi- 
ble aspects of Spacp Station design and program management and, in fact, 
has gone as far as placing a contractual requirement on all contra(,cors 
to use the - stem in specific applications. General applications 
specified include: electronic mail, transmittal of engineering drawings, 
and transfer c all contractually deliverable data requirement list 
items, such as: monthly progress reports, design and trade study 
results, data packages, and cost and performance data. Specific applica- 
tions will be developed by individual organizations, and as their utili- 
ty become apparent, will be made available to all others. Under NASA 
direction, some applications will be mandatory for specific functions in 
support of the program. For example, NALA may dictate that all Review 
Item Dispositions (RIDs) be processed through a specific RID Tracking 
Application activated to support a specific programm design review. 

Generic Applications Potential for TMIS 

Generally, any need for information exchange across organiza- 
tional lines can benefit from the facilities of the TMIS network. During 
the Definition and Preliminary Design activities, there is a continuing 
need for dissemination and access to information from ongoing trade 
studies that affect more than one function or end item. The TMIS can 
fulfill this need oy serving as the storage for evolving requirements 
and interim analysis results. In addition to providing the data and 
storing results, the TMIS also provides the analytical tools to process 
that data and generate information. Technical data may be in the form of 
text, ^laphics, or tables. 

Management functions supported by TMIS include the already men- 
tioned contractual reporting; task planning, scheduling and status 
tracking; cost and schedule performance monitoring; and di.33emination of 
personnel locator and assignment information. 

Specific Applications Now On Line 

At this time, NASA is using the Telemail system for electronic 
mail among centers and contractors. Level B program management is using 
a Cyber 830 computer for budget planning and tracking functions. Con- 
tractors, on the other hand, are implementing their own network for 
information exchange with all members of their respective teams. Table 3 
is a partial list of the applications in use by the McDonnell Douglas 
Team. A large majority of these applications are built using the facili- 
ties of a rt i.i'.ionai database management system (RDBMS). 

In determining which applications to implement, emphasis has been 
placed on the potential productivity gains achievable by the use of each 




application. Concern for productivity is carried through in the actual 
development of the applications and it is this concern that led to the 
selection of an RDBMS for that purpose. Significant productivity gains 
can be achieved by using an RDBMS for developing applications when 
compared to using a language such as PASCAL or COBOL. Gains of five tp 
tenfold have been reported in programming productivity along with claims 
that certain applications would just never have been developed at all 
without availability of the RDBMS (Reference 2], 

Other Potential Specific Applications 

As with any other powerful tool, the value of THIS to the program 
will grow with the number of applications the system supports. The crea- 
tivity of users and system developers will determine hov many appli- 
cations and to what extent TMIS will support them. For example, Engi- 
neering and scientific reference data can be made available on the 
system to ensure consistency and accuracy of the data used by all play- 
ers in the design effort. Using TMIS for storage of this type of refer- 
ence data requires large storage capability and fast searching software 
tools. Technology is evolving to the point that today we have special 
purpose database processors, like the Briton-Lee Tnteliigent Database 
NIachine (IDM), and very powerful dafafjase mfi'-.-^gement software that bring 
the desired capability wirhin reach. 

One of the potentially most useful applications that TMIS may 
support in the future is document management. This application would 
support preparation, review, approval, maintenance, and configuration 
control o<: multi-disciplinary documents such as interface control docu- 
ments (ICDs) and end item specifications. These documents require active 
involvement by many organizations, responsible for the contents of 
different sections or types of information. It is possible for all the 
cognizant, organizations to maintain control of their specific areas 
through proper allocation of security authorities dealing with changes 
and approval for changes, Changes can be coorainated using Change Re- 
quest fiies that can be electronically mailed to those potentially 
affectf.d by the ciiange. Re iews and coordination of changes cai be 
conducced using scratch files end electronic mail. Approval of changes 
can be executed by configuration control managers by the use of special- 
ly assigned "signature" passwords. The key advantages of such a process 
is t.hat all reviewers are assured of working on the same generation of 
the subject document and their comments are available for general review 
by the electronically redlined version of the document. 

TMIS allows the aerospace professional to use computers and 
information networks in a manner not unlike science fiction works like 
"Star Trek" and "2001: A Space Odyssey" indicate. The technology 
available to the Space Station Program will provide the Hardware and 
software needed to derive great productivity gains from TMIS. Only 
imagination and the willingness to store and format information will 
limit the extent and the ways in which the system can benefit the 

■"? 159 



■IwWWi <f*5. >-«>. V\«v.vK\-s.».> 




CO c 

H o 









r- CM 

W ^ lO 




♦* o 

m o 

O c 

_ TJ 

S I 

o o 

2 5 

»- CM CO ^ lO (0 



«t — 
w 0) •»' « 
en c « — 


o I? r k. 
> o — ® 

•—■«-• «« 

♦J « X 0» 

== »- CM 


CO ^ lO 

< O O .'J — 

53 i; T " o» 

si — fc ^- lu 

»- CM Cf> ^' lO (d 



*- CM CO ^ 




li- CO 

♦* c 

U ffl 

M « F 

4^ i. 

C « L. 

i- c o (0 

- O DC « 

» ** — "D 

c itf rt 3 

t o» — 

o — ■«-• o 

♦* •»- O "D 

W C C (S 

=^5 5 


»- CM CO >* 











Characteristics Of TMIS That Enhance Productivity 

No tool can increase productivity by itself. Results come only 
from the use of the tool. Information industry literature and experience 
in implementing information management systems indicate that the three 
key factors that stimulate the use of a tool such as TMIS are, in order 
of importance: 

1) Commitment and support from top management, 

2) Ease of use and availability of training, and 

3) Benefit to the user. 

The Space Station Program has done a commendable job of taking care of 
the most important factor by making a clear commitment to TMIS early in 
the program. It will be up to those implementing the system to make su'-e 
that factors 2 and 3 above are of paramount importance in defining 
system components and applications to be developed. 

Item 2 is a measure of what has become known as "user friendli- 
ness". A dilemma arises here in deciding whether to give the user many 
powerful tools such as application generation languages, or to provide 
menu-driven, preprogrammed applicdtipns. The first approach requires a 
significant "capital" investment in developing the skills of many users 
who then can derive great benefits from the system. The latter approach 
reaches more usi'-s, thereby making the system unquestionably more user 
friendly, but at the cost of a more visible "capital" investment in a 
cadre of application developers. These developers, however, become quite 
adept at taking advantage of the suble capabilities of the system and 
are invariably more efficient in the development and maintenace of 
applications than the user who only programs as a sideline to his or her 
job responsibilities. 

The degree to which TMIS is used will depend ultimately on the 
degree to which people accept the work stations and the role that elec- 
tronic information exchange can play in their day to day work. This 
acceptance is growing and can already be measured by indicators such as: 
use of word processing equipment; presence of terminals and personal 
computers, not only in specially designated terminal rooms but on the 
work desk of the engineer, scientist, or manager; use of database man- 
agem«jnt systems for data entry, query, and report generation; prolifera- 
tion of data communication facilities; and the installation of local 
area networks to support individual facilities or agencies. 

Electronic information exchange should not be the means to do the 
same work with less people. It should be a tool to help utilize the 
cognitive capabilities of humans to a greater extent so that we can do 
more work, and work of greater value, with the same number of people. 
Greater productivity can be achieved in quality as well as in quantity 
by working smarter rather than by simply doing mce work. 





As technology evolves and the size and complexity of systems grow 
in the aerospace industry, the need for accurate, timely dissemination 
of voluminous amounts of data has expanded to an almost unmansgeable 
magnitude. The Space Station Program, comnlted to the use of innovative 
concepts for the attainment of high productivity, has recognized the 
importance of information availability in improving productivity by 
specifying the role of THIS in the Program. THIS, however, is a tool 
that must be used if it is to help bring about the desired productivity 
benefits and, to quote one of the managers interviewed during the MIDAS 
study, "it should be used to do better work, not to come up with better 
excuses as to why the work cannot be done". 










[IJ Hughes Aircraft Company, "R & D Productivity", 2nd Edition, 
Hughes Aircraft Conpany, 1978 

[2] Sobell, Mark G., "Database Management Systems Cut Application 

Development Time", Computer Technology Review, Spring 1985, pp. 23-28 



Gonzalo Montoya is an Electrical Engineer with 23+ years experience 
in the Aerospace Industry priniarlly in the area of Computer System 
Software Engineering. He was instrumental in the definition and 
development of the Management Information Database Automation System 
(MIDAS) for Mission Operations Directorate at JSC in his position as 
manager of the project. He is currently assigned to the Space Station 
Phase 6 Study Team and has been working on the THIS and other issues 
for the Space Station. 

Paul Boldon is a Software Engineer with 5 years experience fn the 
Data Processing in'^ustry. He was an integral part of a team, which 
developed a database system for the University of Texas Medical Branch 
at Gulveston, Texas. He is currently assigned to the Space Station 
Phase B Study team as Deputy Task Leader for TMIS for work package Z. 

'i 164 







James W. Dean, Jr., Gerald I. Susman, and Pamela S. Porter 

Center for the Management of Technological and Organizational Change 
College of Business Administration 
The Pennsylvania State University 


This paper presents an analysis of the process of implementing 
advanced manufacturing technology, based on study in numerous organiza- 
tions. This process is seen as consisting of a series of decisions with 
technical, economic, and political objectives. Frequent decisions 
involve specifications, equipment, resources/organization, and location. 
Problems in implementation are viewed as resulting from tradeoffs among 
the objectives, the tendency of decision makers to emphasize some objec- 
tives at the expense of others, and the propensity of problems to spread i 
from one area to another. Three sets of recommendations, based on this 

analysis, are presented. \ 



In the past few years, a variety of new technologies have become 
available for use by manufacturing firms. These technologies include 
CADCAM, robotics, MRP, CNC, and CIM. Responding to the potential gains 
in productivity, quality, and flsxibility offered oy these technologies, "" :f 

many have included them in their plans for modernizing and auto- \ 

mating their facilities. 

While some of the gains promised by advanced manufacturing tech- i 

nology (AMT) have indeed been realized, firms have often experienced ■ 

substantial problems in their implementation attempts. These problems ! 

are often nc technical per se, but stem from difficulties in managing 
the relatioHinip between the technic?! aspects of automation and other 
organizational considerations. The objective of this paper is to ana- 
lyze the problems that firms typically encounter in implementing AMT, 
and offer some suggestions for resolving them, in hopes of increasing 
the success rate of such attempts. 


'."*iS^ -■'.-' - 

■I .V.Wfc , 1_ 


The paper is divided into three sections: 

1. What does the process of technology implementation 
consist of? 

2. What kinds of problems do companies encounter in 
attempting to Itnplement AMT7 

3. How can these problems be avoided or overcome? 


The process of implementing new technology fundamentally con- 
sists of a series of decisions, whirh take place over a pe.iod of months 
or years. These decisions typically include the selection of t^e tech- 
nology to be used, the identification of the vendor, the product or pro- 
cess where the new technology will be piloted, how quickly the 
technology will be spread to other units, and so on. Each decision is 
constrained to some extent by the decisions that precede It, and 
constrains in turn the decisions that come after It. In this way, the 
universe of technological possibilities is gradually narrowed to one 
system, with a specific set of capabilities, and a particular Implemen- 
tation approach. Thus, the resulting technology is the expression and 
sum total of all of the choices made along the way. 

In order to be successful, decision-makers need to simulta- 
neously consider three objectives: technical, economic, and political. 
The technical objective consists of developing a system that will per- 
form the required task according to specifications. The technology must 
move material, cut metal, or process information, in an effective 
manner. Much of the effort is devoted to design and implementation 
demonsti ating that the new technology can indeed perform as advert<;ed. 

Tech;.:cal success is, however, not the sole criterion of effec- 
tiveness for new technology implementation. Since te'~^''l':3l systems arc 
Imbedded in business organizations, the new technolc ' also achieve 
economic objectives. Th's involves demonstrating 'he new tech- 
nology puts the organization in a better f>'ianc. i :> itio<^ than it 
before. Depending on thr organ i'^?Mor,, this riay b. . vreb? n terms 
of payback period, or internal rt.ce of return. Whne >i)"' -m that 
are pushing automation have rel3;<ed the economic c.i ' h^y are 
almost always present to some extent. 

The third objective to be msi in implementing new technology is 
political. This consists of generating support from all the people in 
the organization, often spread over several departments, whose commit- 
ment Is necessary to make the new technology work. The political 
•j problem exists because the technology is being inserted into an ongoing 

social system, and It will have an impact on that system. Few tech- 
nologies will function effectively wh^jn people are indifferent or 
hostile. Thus, in order for ir(ip1emc-ritation to be a success, the rele- 
vant individuals need to be committed, "o,\ board," and excited. 

It appears that part of the difficulty in technology Implemen- 
tation stems from the fact that there are almost always tradeoffs 


3^ ■.\'*S^ ■■'." ' 


Involved among these three objectives. For example, making a decision 
so as to maximize economic return may incur technical costs, and 
choosing so as to increase ♦■he political acceptability of the technology 
may limit its technical success. T' . presence of these tradeoffs alone, 
however, should not cause major problems for implementation. Making 
tradeoffs such as these are the stock in trade of managers. However, 
the technologies currently being implemented are ^ery new, and most 
managers do not yet have a base of experience on which to rely. Thus, 
it often appears that, in making implementation decisions, tho^e is an 
imbalance in the attention devoted to the three objectives identified 
above. Many decisions suffer from overattep*;ion to one or two of these 
criteria in decision making, and insufficient attention to the other(s). 

By basif'' he decisions on only one or two of the criteria, 
managers often create p. lems on the one(s) ignored. Because of the 
connections among the technical, economic, and social systems, problems 
in one area often create p'-oblems in tlie others. In this way, the 
objective that was emphasized in making the decision may not even be 
achieved. A familiar exampTe of this process would be if a ma lager were 
to attend soleJy to techni al and economic concerns in making a deci- 
sion, by ignoring the ooiitical objectives, the manager might create 
problems of j,er accepUnce. Eventually, the problems could become so 
severe that the users would not want to , make the technology work, and it 
would fail. Technical success would thus not be achieved, even though 
it was emphasized in the decision. 

Research that has been conducted in a number of firms leads to 
the conclusion that this is a common scenario in the implementation of 
AMT. There are recurrent patterns involving systematic over- and 
unde attention to i^iecific criteria in specific types of decisions. 
Thus, attempts to implement AMT often run into trouble. The problems 
are not technical per se, but stem from the interdependence among tech 
nical, economic, and political factois. In the next section, we will 
describe how these dynamics have led to problems in the firms studied. 


There are at least four uifferent ypes of decisions common ico 
new technology implementation: Specifications, Equipment, Resources/ 
Organization, and Location. Our research has identified some patterns 
in how these decisions are made that create problems for the imple- 
menting firms. Each of these decision types will now be examined, along 
with examples from research of how excessive or insufficient attentico 
to the three criteria has created problems in the firms studied. 

Specification Decisions 

The first tyoe ^f decision commoniy maoe in new technology 
implementation involves specifications. Simply put, these decisions 
determine what the new system will do, and what it will not do. In this 
way, what usually bt^gins as a vagi. idea in someone's head is 



.'i .s> v*.:'^N^'-v 



cry.-tallized, put down on paper, and becomes the initial vehicle for 
generating support and commitment ?t the project. The protot.vpical 
situation for specification decisions is vihen political critf a are 
underemphasized, and constituencies that should be consulted are 
neglected. Failure to consider the political ramifications of specifi- 
cation decisions often leads to disillusionment among users and support 
personnel, and ultimately limits "le technical effectiveness of the 

Regretting this sort of decision process, one automatio;i project 
manager mused, "I'd like to tell you that we didn't just design It and 
throw it at them, but there was some of that." The case In point was an 
information system, designed for use by manufacturing managers. V'")ile 
expressions of outright resentment of the system by this group were 
constrained by its support in upper management, there was quite a bit 
less than total e. ..lusiasm, esptcially among first-linf» supervisors. 
In addition, a najor component of the information provided by the system 
was considere'' innecessary by the intended users. 

Ironically, soiretimes problems in specification decisions stem 
from overattent on to the political dimension. It has often been noted 
that the initial list of system functions grows throughout the project, 
sometimes to the point where the original gcal is but one of many. This 
is usually a result of new peoole hec^ming involved in a project, and 
saying, "Couldn t we also make it do...?". While agreeing to these 
requests is often an investment in the support of the individuals 
involved, it may lead to an overwhelming list of system functions, none 
of which are accomplished effectively. Ths is a result of political 
objectives being overemphasized, to the detriment of technical objec- 
tives. To ' ilete the cycle, the lack of technical success in the pro- 
ject often u.. enchants the user community, so the political advantages 
of agreeing to the numerous requests are not realized. 

Equipment Decisions 

The second type of technology implementation decisions concerns 
equip.nent. One decision concerns whether technology will be purchased 
or developed in-house. If it is to De purchased, there Is the question 
of which vendor or vendors to select. Often connected with these deci- 
sions is the e> jct type of technology to be used. Equipment dec'iious 
can usually be broken down into hardware- and software- related. A 
repeated finding is that some firms automate using the absolute lowest- 
cost equipment they can obtain. Often, this equipment does not perform 
as expected, and ;:.ids up costing mo/e in the long run. For e'lample, a 
building supply company decided to put their shop floor control system 
on a minicomputer, so as to avoid the high costs of using the corporate 
mainframe. Unfortunately, the minicomputer could only handle half of 
the company's product lines, and it would have cost a fortune to rewrite 
the software for the mainframe. To date the company has only half of 
their business on the system. 

At first blush, this sounds as if it represents an overemphasis 
on economc criteria, and a resulting sacrifice of technical criteria. 
Put often these scenarios arise because managers Are. reluctant to try to 




Uu-iX,,^. ,'X. 


* s 

convince their superiors of the real cost of what they are trying to 
do. Thus, managers try to avoid a political battle by keeping expen- 
ditures at a point where they do not need approval from higher levels, 
or at least will not raise eyebrows at those levels. Sometimes this is 
achieved by dividing a major purchase into several smaller purchases. 
Given this background, which has been observed in a number of different 
firms, it appears that the technological objectives of equipment deci- 
sions are often undermined by a combination of economic and political 

Resource/Organization Decisions 

The third type of decisions necessary for the implementation of 
new technology concerns resources and organization. The issue here is 
the 2. . -nt of manpower to be devoted to the implementation attempt, and 
how i:.^ individuals who join forces in this effort will be organized. 
It is our observation that the level of manpower devoted to new tech- 
nology implementation is often inadequate. Firms try to conserve funds 
by doing a project with half of one person's time, a quarter of another 
person's, and so v-n. Since people generally feel more loyalty to the 
tasks they have been doing and are familiar with, they often do not even 
devote the amount of time allotted to the implementation project. This 
problem is compounded by the fact that performance appraisal and reward 
system seldom adequately recognizes work of this sort. 

In one instance, an accountant was slotted to spend over 50 
percent of his time on a new MRP system. Unfortunately, the company 
had made no provision for the fact that the fiscal year was ending, and 
no one had been assigned to cover for him. Needless to say, not much of 
the accountant's time was devoted tc the MRP. In another instance, an 
electronics firm was simultaneously implementing an automated material- 
handling system, a shop-floor control system, and moving into a new fac- 
tory. This was attempted with only one full-time person devoted to the 
projects. Numerous technical problems were not able to be solved, 
simply because no one had the time to do it. The whole situation was 
demoralizing to both the implementers and the users of the new tech- 
nology. As one of the foremen said, "In the long run, they'll wish 
they had spent a few more salary dollars." Characteristically, the 
at^'empt to save money by undermanning the project backfired. Because 
there was no funding for technicians, highly paid systems analysts ended 
up doing menial work such as running cables. Thus, the economic objec- 
tives sought by undermanning the project were not really achieved. 

Related to decisions involving the lev^l of human resources 
applied to an implementation project are decisions regarding how these 
individuals will be organized. It has been our observation that, in 
order to successfully implement AMT, individuals from several different 
parts of the organization need to coordinate their efforts. Depending 
on the specific technology, individuals from any of the engineering 
disciplines, production, information systems, (Accounting, and marketing 
may be involved. When, political problems be.tween two or more depart- 
ments serve to hinder coordination, th^^ technical integrity of the pro- 
ject may be compromised. 




- ■*--'#«r^3; 


One example of this is a project which was to develop an auto- 
mated quality data collection system. The project was being championed 
by a plant-level quality group. While H appeared that their efforts 
could have been enhanced by help from the corporate quality and systems 
organizations, political problems among these groups led to the original 
group "going it alone." In another case, the Installation of a CAD 
system wa«- delayed for over two years by coordination difficulties bet- 
ween the product design and manufacturing departments. 

The irony of this situation is that often the political problems 
can't be solved until the new technology exhibits some level of success. 

\ In order to protect their careers, people in organizations try not to be 

associated with anything that fails. So it often appears that potential 

' participants and supporters are "waiting in the wings," to see if this 

i uncertain technology will deliver as promised. If the technology 

falters, people will lose faith in both the system and its proponents, 
any initial enthusiam will be lost, and the chances of political success 

■ with the undecided will become more remote. As those who might poten- 

tially be able to help withdraw from the project, the likelihood of 
technical success decreases in turn. 


Decisions about organization that are technically motivated may 
have outcomes that are quite unpredictable. In one firm, a woman who 
wrote procedures for assembly line workers got involved with the 
installation of a shop-floor control system. Her duties expanded to 
include trouble-shooting with the bar-code scanners used to collect 
data, and training for the system users. While neither her job grade 
nor her pay wer«? increased, her elevated status was resented by the 
other procedure writers, many of whom had greater seniority. This pre- 
sented difficulties for the project, as they would not make time to meet 
with her when the system was being introduced in their areas, making it 
difficult to expand the system beyond the initial test bed. 

A final example of organizational decisions comes from a multi- 
division firm in a metals industry that was attempting to implement CIM. 
When the computer professionals within the firm first suggested that CIM 
was an important strategic direction for the firm, they were virtually 
sent packing, because the senior managers felt that the idea of computer 
integration ran counter to the prevailing corporate spirit of decentra- 
lization. The only way that CIM was eventually approved at the senior 
executive level was that the CIM advocates packaged it as something that 
could be under complete control of the business units. When it came 
time to develop the system, however, this organizational arrangement 
made it extremely difficult to devise a system that could be used 
throughout the corporation. Satisfying political objectives thus under- 
mined the technical goals of the project. 

Location Decisions , 

The final of the four types of decisions involves location. 

This includes thfi selection of an area (plant, product line) in which to 

pilot thfc new technology, as well as other areas in which it will be 

implemented. Politics often dictate which areas are selected, and which 

! are avoided. In one case, the new technology was piloted on Product X. 


As we observed the implementation process, we gradually realized that 
many of the start-up problems were being caused by the complexity of the 
product. When we asked the project manager about this choice, it turned 
out that he and others at his level would have preferred to pilot it 
"anyplace other than [Product X]." This decision had been made at a 
higher level, and was based on the politics of selling the new tech- 
nology to the management of the various products. In a variation on 
this same theme, a company was selecting the product line for a robotics 
application. The line which was selected was primarily for economic 
reasons, because it was experiencing a large order backlog. It turned 
out, however, that the assembly process chosen was extremely compli- 
cated, and led to technical problems in developing the system. In addi- 
tion, the pressure to ship products in this area kept the operators from 
spending time on training, which created further delays. 

Our goal in describing problems that firms have in implementing 
AMT is not to discourage firms from attempting it. On the contrary, we 
hope that awareness of these problems helps firms to avoid them in 
future attempts. In the concluding section of the paper, we offer some 
suggestions for avoiding the kinds of problems that have been 


Our overall recommendation is clearly that, in making the deci- 
sions of which technology implementation consists, one must be aware of 
the impact on all three of the areas. Many problems are caused by one 
area inking undue precedence over the others, or potential negative 
effects 0^ a decision being overlooked. While awareness of the impor- 
tance of e&rh factor does nothing to ameliorate the problem of trade- 
offs, it at ]°ast alerts managers to the potential problems caused by 
Gverattention to one or two of the objectives. In order to materially 
improve the implementation process, howrver, managers need to find ways 
to go beyond the tradeoffs, and attempt to jointly achieve technical, 
economic, and political objectives. This is, of course, more easily 
said than done. In many cases, the pre<:crice of tradeoffs among the 
three factors is simply a dilemma which must be faced. This should give 
one pause in evaluating the potential of "easy," off-the-shelf methods 
of new technology implementation. Recommendations for better practice 
should be grounded in an awareness of the complexity and dilemmas 

An overview of the problems discussed above would revecl the 
following process. In trying to limit expenditures on technology, mana- 
gers buy or develop systems that are not as good as they need to be. 
The effectiveness of the technology is further limited by organizational 
considerations, such as barriers between departments. Over time, the 
technical shortcomings of the system give rise to political dif- 
ficulties, as users and others whose support is needed become progressi- 
vely more disenchanted. This eventually becomes a "vicious circle" of 
declining technical performance and political support. The ultimate 
outcome of this scenario Is economi", as the benefits originally pro- 




mised by the new technology are not realized. Usually, however, before 
the situation reaches this point, management intervenes. These inter- 
ventions usually take two forms: more money and people are thrown at 
the problems, and the project manager is impressed with the implications 
for his or her career if the system is not soon operating as advertised. 
This threat then cascades down through the project organization. 
Simultaneously, the project manager and others are working both up and 
down the hierarchy to lower expectations for system performance. 

As indicated above, there are limits to the extent to which 
these problems can be avoided, given the tradeoffs among objectives, and 
the fact that problems tena to spread from one area to another. 
Discussed below, however, are three sets of recommendations for mana- 
gers, which are derived from the cases studied, and the conceptual 
framework abstracted from them. These recommendations emphasize the 
anticipation of problems before they occur, and the breaking up of the 
feedback loops, in which problems lead to more problems. 

Technology and Resources 

The first recommendation, and the one which will probably take 

the most courage to implement, 
to convince their sfjpei^iors to 
deve":op an effective system, 
and the personnel to properly 
this attempt ^'s underlined by 

is that managers should make every effort 
spend the money necessary to purchase or 
This includes both the technical system 
implement it. The importance of making 
the fact that, in a great number of cases 

studied, probK">ms were initially created by managers trying to save 
money on technology, and these problems escalated into major threats to 
the technology's effectiveness. It is extremely rare for managers to 
feel thai- tliey should have spent less money on a system. This 
suggestion is made particularly difficult to practice 
in many firms, the ultimate decision makers on capital 
have technical backgrounds. Thus, the temptation is 
dations based on cost and return, with less consideration given to the 
long-run technical and political effects of this approach. 

by the fact that, 
investment do not 
to make recommen- 

On the other hand, deciding not to even ask for what is really 
needed, on the assumption that one will be turned down, is a self- 
fulfilling prophesy. Sometimes, managers who take the time and the 
risk to sell their superiors on the need to spend money on technology 
are rewarded. One manager who "decided to make the guy earn his money, 
and tell us why we couldn't have what we wanted," actually got what he 
wanted. It is important to reemphasize that often this approach will 
save money in the long run. Even if the request is not approved, there 
are advantages to making the attempt. If the "champion" has to settle 
for less in terms of technology, if it doesn't work, this individual is 
at least on record as saying that more was needed. Also, making these 
arguments repeatedly may eventually sensitize upper management to the 
general problem, and lead to a more receptive climate for future 

Technical Effectiveness through Organization 


tiveness by 

first recommendation reduces to achieving technical effec- 
spending money. Technical success can also, however, be 




i te k.^i'c. 


enhanced by organizational arrangements which do not have an economic or 

I political cost. For example: in picking an area or product in which to 

. -; introduce advanced technology, it is important to take advantage or 

f existing organizational strengths, and avoid organizationally weak 

\ areas. Implementing new techi><>logy will tend to amplify both strengths 

\ and weaknesses. The ideal area for introducing new technology would 

' have, for example, good leadership and coheslveness. 

Since several different subunits will probably be involved, both 

'\ in the project organization and on the user side, individuals from these 

'\ groups work together well should be selected for the project team. 

Recent work by the authors [1] has suggested that individuals who have 
^ complementary reward structures, broad backgrounds and knowledge, and 

are adept at sharing information and taking another's perspective can 
have a positive impact on the success of implementation. It is also 
.' helpful to have peoph; involved who are willing to experiment, and who 

'{ are not overly tied to the old ways of doing things. Some flexibility 

\ ill terms of going beyond job descriptions will probably be necessary, so 

J individuals who are willing to go in this direction (and those who 

•| aren't) should be identified. If such individuals are not available, 

J some team-building activities may be helpful. 

J By carefully arranging the organization of a project, managers 

^ can improve the technical success of the new technology without 

'l absorbing economic or political costs. Thus, organization provides a 

t way to avoid the vicious circle of technical and political failure. 

Tolerance for Failure 

Developing tolerance for failure is another way to disconnect 
technical problems from political problems. The goal of this effort is 
for users and support personnel to redouble their efforts when ' 

confronted with technical problems, rather than getting disgusted and 
distancing themslves from the project. To achieve this goal, it is 
important for firms to be somewhat conservative in their initial automa- > 
tion attempts. It is better to pick an application that has a high '' -J 
probability of success than one that has a high expected return. "^j 

Developing momentum for automation within a firm is easier to do so with ! 

a few small successes than with a big failure. In one firm, the failure j 

of an ambitious robotics project led to negative feelings about robots \ 

thdt lasted several years. It is crucial that the first attempts a firm I 

makes at AMT do not leave people with a bad taste in their mouths. j 

Otherwise, the first sign of -technical problems in subsequent efforts 
will generate an instant political reaction. 

The most time-honored solution to political problems is to 
involve users in the decisions that create the system. User involvement 
should also help to break the cycle of technical and political problems. 
While most managers are probably aware of the need to do this, time 
constraints and reluctance to negotiate with factory floor personnel 
over technical matters keep them from actually doing it. It is impor- 
tant to keep in mind that there is an inverse relationship between the 
amount of time it takes to make a decision and the amount of time it 
takes to implement it: the time that is not put into user involvement 
on the front end will be spent on the back end. ? 



As the new system develops, the establishment of regular com- 
munication links to keep people up to speed is essential. Early in the 
project, it should be impressed on all participants that few tech- 
nologies work right immediately, and that some problems should be 
expected. Eventually, this may include reminding people of start-up 
problems in systems that were ultimately a success. As the project 
progresses, communication needs to be maintained. 

While progress reports via memos are the absolute minimum, 
periodic meetings among Involved participants are better. Meetings have 
several advantages: First, they limit the extent to which technical 
difficulties become political problems, by supplanting the organiza- 
tional runor mill in providing information about the technology's level 
of success. The absence of hard information seems to leave the door 
open for "doomstay" rumors to proliferate. Periodic meetings devoted to 
sharing information should diminish the likelihood of this occurring. 

Additional advantages of frequent project meetings include the 
fact that they allow the "accidental" solution of numerous little 
problems that arise in the course of implementation. We have often 
observed the time immediately before and after meetings being used by 
pairs of individuals to solve problems of this type. Also, working 
together in meetings helps to mold the- individuals involved into a team, 
so that their future potential to work together is enhanced. 

To minimize the time devoted to these meetings, we would suggest 
that all those Involved in or affected by the project be identified. 
These people should be kept informed by memos of progress on the imple- 
mentation. Periodic meetings should include only those individuals who 
are involved in the Issues to be discussed. The project manager should 
have the power to ensure that the right people will be present at the 
right time. (Of course, if he or she has to work very hard to do this, 
there are probably fundamental political problems that need to be dealt 

Our final recommendation, and perhaps the most Important, is to 
attempt tc inspire the people who need to make the system work. This 
more than anything else, should prevent the political problems that 
usually follow technical difficulty. This can be accomplished by giving 
people a vision to work toward: a picture in their mind of where the 
project is going, and a burning desire to get there. Often, the 
"champions" of a new technology have such a vision, but they fail to 
communicate it to the troops. Many possibilities exist for com- 
municating the vision, including site visits to other firms where it 
works, and videotapes demonstrating the concept. A key component is 
confidence in the face of adversity. As an engineering manager who had 
successfully Implemented a CAOCAM system said, "It's not sold in the 
meetings.. .you get a lot of this stuff accomplished by being one hundred 
percent positive in the Informal contacts." If the vision of the cham- 
pions can be shared with all of those whose dedication is necessary, the 
chances are much improved of creating systems that will be successful 
technically, economically, and politically. 






[1] Susman, G. I., J. W. Dean, Jr., and P. S. Porter, Departmental 
Interfaces in the Implementation of Advanced Manufactur*ing 
Technology, working paper, Center for the Management of 
Technological and Organizational Change, College of Business 
Administration, The Pennsylvania State University, 1985. 


James W. Dean, Jr. is Assistant Professor of Organizational 
Behavior in the College of Business Administration, The 
Pennsylvania State University. His interests include the implemen- 
tation of advanced technology, decision making, •■»nd organizational 
innovation and change. 

Gerald I. Susman is Professor of Organizational Behavior in the 
College of Business Administaticr, The Pennsylvania State 
University. His research involves sociotechnical systems, new 
technology implementation, an-d organizational change and 

Pamela S. Porter is a Doctoral Candidate in the College of Business 
Administration, The Pennsylvania State University. Her interests 
center around the design and implementation of advanced technology, 
and how this is affected by organizational structure and culture. 


.^ __^® 




F. T. Crucian, The MITRE Corporation, Houston, Texas 


In determining the productivity of a computing 
capacity, two of the costs to be considered are: the cost of 
lost productivity (user lost time) due to inadequate computer 
resources; and the cost of increasing the capacity of the 
computer system to reduce user's" lost time. This document 
presents the results of a study conducted at NASA/JSC. The 
purpose of the study was to relate the cost of users' lost 
time to the cost of increased computer resources. The goal of 
the study was to identify the overall least cost to the 
computing facility. The document describes a survey designed 
to identify the user's lost time and the computer resource 
require'iant to reduce lost time. The results of the survey 
are preser.ted showing the trade off between user's lost time 
and coct cf increasing system capacity. 


For the past several years the MITRE Corporation has 
provided support to the Central Computing Facility (CCF) of 
the Data Processing Systems Division (DPSD) at NASA's Johnson 
Space Center (JSC). This support has been in several areas 
including the gathering of requirements and the long range 
planning and procurement of additional computer systems. In 
providing this support, MITRE developed a method for comparing 
computer systems costs to costs incurred by the users due to 
insufficient computer resources. This paper describes the 
method used to accomplish that comparison. 




2.0 Background and Approach 

I 2.1 General Discussion 

i The government procurement process req\iires 

I justification for new or replacement computer systecs. The 

justification is prepared by the originating organization and 

', submitted to higher level management for approval. It 

includes a description of the system configuration, a 

statement of requirements, the cost to the government to 

procure the computer system, and the benefits that would 

accrue if the system were acquired. In the approval process, 

f the request la frequently returned with a request for more 

[I definitive statements of requirements and costs-benefits 

'I analysis. Frequently the user's requirements are prepared 

. i from data collected on a yearly basis from first and second 

<i level management. Data from this source can lack 

"granularity" because the resource requirements are usually 

raw requirements - i.e., requests for computer hours, mass 

storage, etc. The end user is not asked to describe how the 

task will be accomplished. Consequently, a method is required 


1) determtr.i che adequacy of the current computer resources; 

2) (^trline the requirements needed to satisfy the user; and 
"^^ compare the costs of the required computer resources to 

the user's costs. The goal is to identify the realistic 
costs of the requirements and compare these to the 
savings and/or benefits accrued to the facility. 

2.2 ^ser Cost vs. System Cost 

MITRE has proposed a method for determining the total 
cost of computer services to an organization. This method is 
based on th* relationship of user response time to the amount 
of system resources. In this relationship the total cost 
includes both the cost of system capacity and the users. (See 
Figure 2.2-1). The plot of Figure 2.2-1 shows three separate 
curves: a user cost; a systems cosc; and a total cost. The 
user cost curve reflects the opinion that the cost of user 
lost time due to insufficient computer capacity will be 
reduced as additional capacity is provided and computer 
services improve. However, as systems capacity increases, the 
systems costs also increase. This relationship is reflected 
■"' in the systems cost curve of the figure. These two "curves" 
^ are summed to produce the total cost curve. This curve 
includes the total costs to the facility for both user and the 


:'*E5^v- . 


^i^nVMfc-Vwku -w - •^«< 


computer resources and also reflects the otal cost/ job to the 
facility. The left aide of the "total cost" curve reflects an 
overloaded system and a cost/ job high in user lost time. The 
right side of the curve reflects an under-utilized system and 
a cost/ job high in systems cost. 

The equation of the user cost curve is not known Ljt 
is assumed to be a curve and not a straight line. Thi-i can be 
assumed because, as additional capacity is provided, service 
to the users improves and lost time is reduced. However, the 
first increments of increased capacity are more effective in 
reducing lost time than the last increments of increased 
capacity. For example, consider users w^th almo':t zero 
computer capacity being required to do their computational 
jobs manually. User cost would be quite high. Adding modest 
amounts of computer capacity decreases these costs 
significantly, and after a certain point is reached, only 
minimal gains are realized. 

Figure 2.2-1 




The system coat curve shows how system costs change as 
system capacities increases. The system costs could be a 
straight line depending on pricing policy of the vendors and 
the size of the capacity increments. For example, a systems 
configuration consisting only of microprocessors would produce 
almost a straight line. It should be pointed out that an 


.-^^^.qjjr-v- / 

ij,^;^^3^>^\ ^^-:^<-'-^ -■ (^ 

almost straight line cost has more of an effect on the total 
cost ci rve than a cost curve similar to the cost curve in Fig. 
2.2-1. The resultant cost curve reflects the sum of the 
changes in both user and system costs The left part of the 
curve shows that the cost to the facility is high in user 
costs if the computer resources are overloaded. The right 
part of the curve shows that the costs to the facility are 
high in systems rosts if the computer resource,"? are 
underutilized. Tl\is curve will reach a low point which 
represents the minimum cost to che facility of coth the users 
and systems costs. 

2. 3 Approach 

The system cost curve In Figure 2.2 1 can be easily 
derived by costing the ranges of capacity. How'iver, the user 
cost curve can only be determined and e/aluated from 
information provided by the user. The user must provide the 
amount of lost time experienced for the current computer 
resources and project the amount of computer rerources needed 
to essentially eliminate the lost time. Frrm thi.s data it is 
then possible to attempt to draw the two curves in the figure. 
The User Cost Curve can havs three poi'. s : 

l)a point on the Y axis corrrsponding to "zero" comput 
er capacity and the total cost of r'le users 

2)a point when the X valu*" is the size of the current 
configuration and the Y value xs the cost of the lost time 
using the current conf iguraticn; and 

3)a point where tb'. X value is the size of the project 
ed configuration and the Y value is the cost of the projected 
system to reduce lost time costs to zero (or close to zero). 

The total cost curve was then plotted by summing the V 
coordinate values for each X coordinate to form the X and Y 
coordinates for the points on the to'-al cost curve. 

The methodology used was based on data acquired via 
questionnaires provided to the users of the computer facility. 
The objective was to identify the time ?.ost because of 
insufficient computer resources and to state the computer 
resources needed to eliminate lost. time. Statistical 
processing of the data allowed the results of the 
questionnaires to be viewed in a variety of ways. For the 
purpose of the survey, lost time was defined to be all idle 
time caused by insufficient computer resource." and include" 
time spent inefficiently as well. This is discussed in more 
detail later. 




'.'*^- •'.-' . 

?J riNWvA .%.-.' ?■>. '!""::/-'? >^ 

2.3.1 Description of Ouastlonnalre 

A two part questionnaire was designed to identify the 
information. The first part of the questionnaire pertained to 
the user and the organization. The questionnaire included 
questions that identified organizational information, the 
individual identification data, resource usage, and lost time 
data. Information about the organization included location 
aata, task assignment, engineer Ing packages required, etc. 
The individual identification data Included specific run data, 
account numbers, physical location ot the user, experience 
level, etc. Actual resource usage included estimates for 
total resource (SUP) usage. Batch and Demand (interactive) 
us?, and terminal usage. The lost time data included time 
lost waiting for output, using demand terminals (slow 
response, facility waits, etc.), because of hardware failures, 

The second part of the questionnaire allowed the user 
to record, the estimates of resources required to eliminate 
lost time and how the resources would be used. In this 
survey, the user was asked to specify computer resource 
requirements, expected turn-around time, and the time of aay 
the batch runs were submitted or for the demand terminal 
session. Data provided in this form were used to form a 
profile of computer usage by time of day. The profile was 
then usci to identify peak requirements, prime time averages, 
overnight requirements, e':c. thus describing the computer 
resource requirement in finer detail than a total requirement. 
Once the profile was determined then it was possible to select 
the "requirement" to be met — e.g., two hour peak load, prime 
time average, etc. Once the requirement was selected, it was 
possible to identify the computer configuration needed to 
satisfy that requirement. 

In preparing the questionnaire, the specific wording 
of the questions was considered to be very important, A draft 
was prepared and reviewed by a small group whose primary 
Junction was capacity urd configuration planning. Tne review 
process required many iterations before the questionnaire was 
considered satisfactory. Questions had to be self explanatory 
and words had to be selected that were not ambiguous. It was 
also necessary to prefix a short memo to the questionnaire 
defining the meaning of "lost t-'.me". It was critical that 
lost time be defined uniformly -- not as a large number of 
different user? groups might define it. (To illustrate, ask 

1 SUP - Standard Unit of Processing. The SPERRY 1100 measure 
of computer resource (CPU, ID, etc) utilization 



several different people to define lost time. The definition 
will be forthcoming only after a few minutes of discussion 
with each person.) The questionnaire defined lost time as 
that time spent waiting for: return of output; retrieval of 
files and tapes; slow response on terminals; inefficient 
practices caused by resource limitations (e.g. insufficient 
core ^nd SUP allocations, etc.); hardware failures; lost 
output; extra travel to retrieve output from other areas; 
travel to other areas to use graphics terminals not available 
at the place of work; and a special "other" category to be 
defined by the user. Time was not considered to be lost if 
the user had other duties that did not require computer 
resources — e.g. attending, reading documents, etc. 


3. 1 General Discussion 

The 3urvey was performed for an entire division of the 
engineering users of the Central Computing Facility. This 
division consisted of five separate organizations: a civil 
service group involved in the monitoring of contractor 
activities and development of engineering programs (Group A) , 
and four coiitractor organizaticns. The four contractor groups 
each had separate but supporting functions. Group B is the 
largest of the four contractors and is involved in engineering 
tasks using computer system and software packages, with the 
primary responsibility of mission planning. Group C is 
responsible for providing software maintenance and 
enhancements for all the engineering processes developed for 
Groups A and B. Group D consists of software specialists 
involved in the maintenance and enhancement of general 
software packages used by all other groups. Finally, group D 
is responsible for providing general support to Group A. 

At the time of the survey, the five groups used two 
Sperry systems - an 1100/81 and an 1110. These systems 
supported 100 concurrently active demand terminals which 
included 30 orraphics terminals. The user could direct the 
computer output to a variety of devices - Sperry 770 line 
printers, Datagraphix Mini-Auto COM Fiche printer. III FR80 
an CALCOMP plotters, and remote printers. Group D was the 
only group that did not have a remote printer or graphics 
terminals at their work area. During the prime shift (0730 - 
1700), demand runs were given a high priority. Runs with 
large memory requirements were carefully monitored. A lew 
batch runs were allowed to be submitted during the day but 
most batch runs were processed after prime shift. A backlog 


■; ^'"S^-.' 




of unopened batch runs would build steadily during the week 
and would be worked off over weekends. 

3.2 Survey Results 

Approximately 700 questionnaires were distributed to 
the five groups. Of the 600 surveys returned.- 320 of the 
responders declared themselves to be users of the CCF computer 
systems. Figure 3.2-1 presents a sample of the results of the 
analysis of the first questionnaire. This figure shows the 
results according to each group and the hours of tim? lost 
each week because of insufficient computer capacity. The time 
lost was attributed to the following reasons - waiL for 
output, wait for terminal facilities (including poor 
response), wait for file access, insufficient allocations, and 
Hardware Failures and Reruns. The category "other" was folded 
into the last category. The time lost is also repeated as a 
sura for each group for the survey week and as a total hours 

FIGURE 3.2-1 

1 SHPb. 1 FOR 1 rOR 1 rOR lALLCC. 1 RCRUHS > AVG. 1 HOURS 
1 lOOTfrt 1 TtR«. 1 flLM 1 1 1 "«!<• 1 

MX 1 )]0 1 0.» 1 l.« 1 0.» 1 1.$ 1 O.J 1 4.7 1 IS04 

i 1 1 1 1 1 II 

OROOP A 1 4] 1 O.J 1 I. J 1 O.J 1 i.l 1 C.J 1 J. 7 1 IS* 

1 1 1 1 1 1 II 

CROUP ■ 1 Its 1 O.t 1 1.4 1 O.S 1 l.f 1 0.4 1 S.O t 110 

1 1 1 1 1 1 1 [ 

"aiwir'c ?"49 ""orrT'i.i ? o.s I i.o 1 0.7 1 4.j 1 jii 

1 1 1 1 1 1 > 1 

'oROOP 1 l» j 1.4 1 J.J '\ 0.5 1 0.» i 0.4 j 5.S 1 4 

CROUP E 1 IS 1 O.B 1 1.5 1 O.J 1 0.9 1 O.l 1 J.S 1 5J 
1 1 1 1 i 1 ■ * 





As can be seen, all users lost a total of 1504 hours 
during the week for an average of 4.7 hours /user /week. Group 
D had the highest average of lost time mainly because of the 
time lost waiting for output and terminal facilities. This is 
to be expected because this group did not have access to a 





remote printer and traveled to different locations to use 
graphics terminals. Group B had the next highest average and 
the largest total of hours lost. Since this was the largest 
group - 61% of the total sample, the overall average was 
affected strongly by this group. Group B reported nearly 
twice as much time lost because of insufficient resources as 
any other group. All other categories of lost cime for Group 
B were fairly close to the other groups. Group C reported 
more lost time due to hardware failures /re- runs than other 
groups, and waiting for files. (Further analysis would be 
necessary to identify the reasons for this.) The cost for 
these 1504 total hours of lost time was calculated at a rate 
of $30/hour/user and amounted to $45,120/week or 
$2,346,240/year. (The $30/hour rate is a figure that reflects 
the total costs to the government and includes such things as 
vacation, sick time, benefits, contractor overhead, profit, 
etc . ) 

The second questionnaire asked the user to identify 
the batch run and demand terminal session resources required 
to eliminate all lost time. Each c-f the 320 users was asked 
to collect the required runs into groups and specify five 
separate variables fcr each groups oi runs - the number of 
runs, time of day the group of runs would be submitted, memory 
size of the group of runs, average SUP minutes required for 
the group, and the turn-around time needed for the runs. In 
the case of demand runs, the turn-around time was changed to 
demand terminal connect time. Figures 3.2-2 r.nd 3.2-3 show 
the Computer Resources and Demand Terminal Resource profiles 
derived from the questionnaire. 

Figure 3.2-2 shows the profile of the SUP requirements 
as described by the users. This figure was drawn by combining 
the weekly SUP requirements for 12-two hour periods (X axis). 

of 0800-1000 hours is the 

period of OBuO-1000 hours 

gives the SUP hours /hour 

According to the second 

for the week was 1798 SUP 

Thus, the requirement for the period 
sum of all the requirement for the 
foi the total week. The Y axis 
needed to satisfy the requirement, 
questionnaire, the total requirement 
hours. Assuming 15 shifts of operations (120 hours), this 
would require a configuration delivering an average of 15 SUP 
hour 3 /hour. However, a peak requirement of 23.5 SUP 
hours/hour was identified occurring between 1000 and 1200 
nours. Two other averages are: the 8 hour prime time average 
(0800 - 1600 hours) of 20.7 SUP hours /hour; and the 10 hour 
prime time average (0800-1800 hours) of 18.5 SUP hours/hour. 
Obviously, the user feels that most of the computer usage must 
be provided during the daytime hours if lost time is to be 
eliminated. It is also interesting to note that about 50% of 
the daytime requirement is for batch runs. 




wi .y^:ri 


FIGURE 3.2-2 


8 HR. 
10 HR. 









Figure 3.2-3 shows the demand terminal requirement 
profile. This figure was drawn from data provided as 
described in the previous paragraph. This figure shows the 
numher of Concurrently Active (CA) terminals over 12 two hour 
periods of the typical day as described by the users. The 
figure shows a peak requirement of 204 concurrently active 
terminals for the time period from 0800-1000 hours. The 
average of concurrently active (CAA) terminals for the prime 
tiine period of 0800-1600 hours is 167 terminals. 

Figure 3.2-4 is a summary of the Computer Resource and 
Der.iand Terminal Requirements. The projection base provides 
four periods for the week used to determine the following 
requirements: the peak two hour period; the 8 hour prime time 
average; the 10 hours prime time average; and the average for 
15 shifts of operation. The figure provides the delivery rate 
in SUP hours/hour, the total capacity in 120 hours of 
operation if suiting from the rate, the Concurrently Active 
Avenge 'CAA) cf terminals, the number of terminals required, 
and an e ::iraate of the number of 1100/81 's required to provide 
the capacity. (It has been determined that an 1100/81 
wil' deliver 2.7 SUP hours/hour when processing the case 
\io ■cload described earlier.) As can be seen, the total SUP's 







available from eeich requirement ranges downward fron 2820 to 
1800 SUP hours for the four projection bases. This was 
determined by <ij«sumlng the stated rate for 120 hours of 
operations. The number of terminals required to support the 
estimated number of CM terminals is set at one terminal for 
each user. The estimates of the number of 1100/81 's required 
ranges downward from 8.8 to 5.6 1100/81's. However, the 
profile (Fig. 3.2-2) indicated that a configuration of 8.8 
equivalen : 1100/81's was required to satisfy the user's prime 
time projection cf resources needed to eliminate all lost 

FIGURE 3.2-3 


210 . 

vwi tn* 

Concurrently Active (CA) 
T^rmi n;il ^ 

189 . 



158 . 

__ _. __ y' 1C7 /DDTMC TTUr 

V 10/ IrKint 1 IMt 

147 , 

l2f. , 

105 . 








1 • 1 • 1 < T 1 T 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 

) 4 8 12 IC 20 24 



f Oay 















23. S 

FIGURE 3.2-4 

1 tPtik Rtq.) 

i8.5,UO_Hr Ay^) 

OveriH Ay£. 












3.3 User Costs 

As a result of the analysis described in the previous 
paragraphs, the points on the user cost curve discussed in 
Section 2.0 can now be determined. The analysis has provided 
the lost time costs of $2.3M using the configuration of an 
UNIVAC 1110 2X2 and 1100/81. At the time of the survey, these 
systems provided approximately 720 SUP hours /week. Using a 
2.7 SUP hours/hour rate for an 1100/81 and 120 hours of 
operations, this translates to a capacity of 2.2 1100/81 's. 
Thus, the user cost curve includes the points where the X 
coordinate is 2.2 equivalent 1100/81 's and the Y coordinate 
reflects a cost of $2.3M/year in lost time. 

Since, it is necessary to provide 8.8 1100/81 's to 
eliminate lost time, the user cost curve must include the 
point where the X coordinate is 8.8 and the Y coordinate 
reflects a reduction of $2.3M/year in user lost time. Figure 
3.3-1 depicts the user cost curve using these points and 
assuming the general shape of the user lost time curve 
discussed in Section 2.0. Using the same approach a systems 
cost curve is drawn showing the increase in systems costs to 
the facility to provide the required configuration to reduce 
user lost time costs to zero. Thi3 curve (or line) must 


? i 

IT t 




IP ^* 


■"., . ,^^ 

include the following two points. The first point has an X 
coordinate of 2.2 equivalent 1100/81 's and the Y coordinate of 
zero increased systems costs for the current system. The 
second point has an X coordinate of 8.8 equivalent 1100/81 's 
and a Y coordinate $5.4M/year for the increased costs to 
support the systems capacity to eliminate user lost time. 
(The "sy£,wem cost curve" is drawn as a straight line. This 
reflects the cost of the 1100/90 family and is affected by the 
vendor's pricing policy. This policy tends to have more of a 
straight line then a curved line relationship. ) Using these 
two curves, a total cost curve was then drawn showing the sum 
of the "delta" costs for both the user and system. In 
preparing this figure the total cost to the facility for both 
users and system capacity was assumed to be the basis for the 
X axis. 

The total cost curve in Fig. 3.3-1 has two X 
coordinates at the V coordinate where the cost to the facility 
is $2.3M/year. On the left side of the curve, the system is 
overloaded and the cost to the facility is $2.3M/year for user 
lost time. This occurs at an X coordinate of 2.2 eqi;ivalent 
1100/81 's. On the right side of the curve, the systeio is not 
as overloaded and the cost to the facility is the sum of 
approximately $1M in user lost time costs and $1.3M in 
increased systems costs. This occurs at a X coordinate of 
about 4.5 equivalent 1100/81' s. The minimum cost to the 
facility occurs at the X coordinate of about 3.5 equivalent 
1100/81 's. At this point, the total cost to the facility 
would total about $2M in both user lost time costs and 
increased systems capacity. Based on the total cost curve, 
the requirements and costs to the facility can now be 
quantified. It can be shown that the facility can: either 
retain the current configuration at a cost of $2.3M/year and 
continue to lose 1504 hours /week of user lost time; or can 
choose to provide a capacity of 8.8 equivalent 1100/81 's at a 
cost $5.4M/year. There are also several choices in between. 
At the same cost of $2.3M/year the facility can choose to 
provide 4.5 equivalent 1100/81 's at an increased cost of 
$1.3M/year and lost time costs of $lM/year. The minimum cos" 
to the facility appears to occur at about 3.5 equivalent 
1100/81 's. At this point on the curves, the delta systems 
cost is about $0.aM/year and the user's lost time cost is 
about $1.2M/year and the total cost is about $2.0M/year. 


-. '.-^Sg^' 



\ '•; 


FIGURE 3.3-1 


3.0 4.0 S.O '0 


Based on the systems costs curve, the facility can 
easily justify a configuration of 4.5 equivalent 1100/81 's for 
the same cost as the current 2.2 equivalent 1100/81 'a. 
However, further justilication not noted here would be 
necessary to justify the 8.8 equivalent 1100/81 's to eliminate 
user lost time. 


MITRE made 
cost/job: (1) is 
overloaded facility; 
underutilized facility, 
showed that as computer 
lost time decrease. In 
identify the total cost 

a presentation demonstrating that the 
high in users lost time costs in an 
and (2) is high in systems costs in an 
A simple method was offered that 
resources increase, computer user's 
this relationship, it is possible to 
to a facility in the combination of 
increased systems cost and decreased users' lost time costs. 
To form this relationship it was necessary to identify: the 
cost of the user's lost time with the current computer 
configuration; the computer resources needed to eliminate lost 
time; and the cost of these computer resources. 




A survey was performed on a selected division of 
NASA/JSC. Two questionnaires were sent to all members of this 
division. Three hundred twenty individuals indicated that 
they were users of the DPSD Central Computing Facility of 4 
SPERRY 1100 computer systems. These users estimated an 
average loss of 4.7 hours/week. The total cost for this lost 
time was estimated at $2.3M/year. The same users estimated a 
requirement of 1798 SUP hours/week to eliminate lost time and 
an average of 167 concurrently active terminals. However, a 
profile was drawn showing that the users needed these 
resources primarily during the prime time. The profile 
indicated a peak use from 1000-1200 hours requiring a capacity 
of 2820 SUP hours (8.8 1100/81 's). This capacity ranged 
downward to 1800 SUP hours of capacity (5.6 1100/81's). The 
profile also demonstrated a peak requirement of 204 
concurrently active terminals. Using the survey information, 
a total cost curve was formed. 

The total cost curve showed that at a cost of 
$2.3M/year, the facility could: continue the cut rent system 
configuration of 2.2 1100/81's and lose the 4.7 
hours/week/user; or increase the systems capacity to 4.5 
1100/81's (an increased cost of $1.3M in system cost and 
decrease in user's lost time cost to $l.M/year). The total 
cost curve also showed that it is not cost effective to 
eliminate all users lost time (a savings of about $2.3M/yr.) 
since the cost to provide 8.8 equivalent 1100/81's is 
prohibitive (about $5.4M/yr.). 





Mr. Francis T. Crucian is a Member of the Technical 
Staff of the MITRE Corporation with 25 years of computer 
system experience. He has provided support to the Johnson 
Space Center for over 13 years as a computer systems analyst. 
He has performed a variety of tasks to identify the resources 
needed to Increase the productivity of the engineering user at 
JSC. Mr. Crucian is the author and co-author of numerous 
technical reports. 


j^ :A.iMv:%vK ^l:^^ 

- V I 







Georgia H. Brock 
Computer Services Division 
Center Support Operations Directorate 
National Aeronautics and Space Administration 
John F, Kennedy Space Center, FL 32899 


In parallel to the evolution of Management Information Systems 
from simple data files to complex data bases, the stand-alone compater 
systems have been migrating toward fully integrated systems serving the 
work force. The next major productivity gain may very well be to make 
these highly sophisticated working level Data Base Management Systems 
(DBMS) serve all levels of management with reports of varying levels of 
detail. Most attempts by the DBMS development organization to provide 
useful Information to management seem to bog down in the quagmire of 
competing working level requirements. Most large DBMS development 
organizations possess three to five year backlogs. Perhaps Office 
Automation is the vehicle that brings to pass the Management Information 
System that really serves management. A good office automation system 
manned by a team of facilitators seeking opportunities to serve end- 
users could go a long way toward defining a DBMS that serves management. 

This paper will briefly discuss the problems of the DBMS organi- 
zation, alternative approaches to solving some of the major problems, p 
debate about problems that may have no solution, and linally how office 
automation fits into the development of the Manager's Management Infor- 
mation System. 


Office automation has many facets, but the rise in administra- 
tive costs has forced industry to seek more aggressive ways of increas- 
ing administrative productivity just as has been done for decades on the 
assembly line. Of course, office work is not a well defined integrated 
process with measurable raw material and countable units of output. 
Therefore, the office productivity axiom assumes that if each office 
task can be completed faster and with more accurate information, then 
the composite of all the tasks will result in greater overall 




k^ . ^ .■V . . ,^ 



productivity. Even haraer to neasure are th» real benefits such as 
Increased profitability or reduced or avoided expenses. At NASA, 
productivity is measured in terms of more work done by fewer people, but 
the amount of work is hard to measure. Increasing launch rates are 
measurable, but the work Involved in new space station challenges is 
f hard to compute or even estimate. Even so, it seems logical to assume 

I that an integrated office environment will produce efficiencies similar 

t to the integrated assembly line. The task of automating the office In 

f itself has potential for increasing efficiency, but every facet must be 

r carefully considered to obtain maximum benefit without disruption and to 

•I create an atmosphere conducive to the process of favorable change. 

X; Since organizations and people tend to resist changes that create 

f confusion antl chaos in the work place, a highly structured evolutionary 

process uust be projected. Office automation must harmonize for the 
benefit of the organization through increased productivity in the total 
management information system. 


and the 


The Dynamic E volution £f the Large DBMS 

It is well known that even the first computers performed simple 
repetitive tasks effectively. Any process that must be done over and 
over by the same identical method is an excellent candidate for 
computerization. Equally important is the computer's efficient ftorage 
and recall o'' data. Once stored, information can be retrieved, sorted, 
and reported to highlight important trends that would have been lost in 
most manual systems. Processing data can be a complicated matheinatical 
model or a simple procedure that manages data to support an organizatioi! 
performing a job. The computerized mathematical algorithm is rather 
easy to Imagine, but the simple procedure in support of a job can be 
clarified by example. For instance, the job of perforraLng maintenance 
on computer hardware seems routine enough for an example of a simple 
procedure. The basic information is the problem report number, the work 
order number, the descr:'ption, and the Identification of the hardware 
component or part. Adding dates provides a history of work performed 
for the maintenance technician, performance information for the 
maintenance technician's supervisor, and identifies resolved problems 
and design changes for operations and design personnel. If the 
organization is relatively large and there are many computers operating 
in similar configurations, (e.g., the consoles supporting Che STS 
subsystems in KSC's firing rooms), then the technician must be identi- 
fied and the location of the hardware established. The operations 
personnel want timely data, so the simple computerized procedure 
the on-line "Autdir -ed Line Replaceable Unit Tracking System." I., pow 
keeps track o. the location of all spare parts, parts sent to the vendor 
for repair anf* expected due dates, etc., etc., etc. It automatically 
flags the on-line "Problem Reporting and Corrective Action System" when 
problem reports are closed. It automatically flags the on-line 


■A T'v'^k nI-v 

"Configuration Management Data System" when design changes are complete. 
It automatically flags the "Shuttle Inventory Management System" when 
the stock of spare computer parts is low. It interfaces with the 
"Automated Ground Operations Scheduling System" to schedule the work and 
the needed resources. Two of the systems that are notified of signifi- 
cant events are not on the same computer. The simple procedure has 
quickly grown into a sophisticated integrated networked system of DBMS's 
that krep track of hundreds of pieces of information that are entered 
by people in different NAS\ and contractor organizations and are 
protected by elaborate securlcy procedures that ensure autonomy for the 
authorized org£.nization. Since these computer pvo^rams essentially fol- 
low the flow of procedures defined to perform work, they are directly 
effected by each change to the procedure. Even adding volume with no 
logical change can affect the computer programs. The comollcated 
niatheniatlcal model is beginning to look simple and the simple procedure 
is bf ginning to look complicated. 

The Prohlams T' at Resis t Solutions 

What is the simple solution to large DBMS that cannot keep up 
with the dynamic nature of woik flow procedures? Can the work flow 
procedures be made less dynamic? Can the compi'ter resources be 
increased to accomplish more timely modifications? Both approaches are 
valid but are not simple or easy in a large organization. 

First, examine the approach that controls the dyncmic nature of 
work flow procedures. KSC has accomplished a major milestone along this 
path by combining a large nurrber of small contracts into two large 
(.ontracts for the base operations and the STS processing. A third large 
contract will handle cargo processing. The model of computer mainte- 
nance in the firing rooms involves the first two major contractors. By 
reducing the number of contractors, the work flow procedures are 
significantly diminished. When responsibilities> are concentrated from 
five or six contractors to one contractor, the computer program becomes 
simpler. However, it must be changed. Along with the scramble to 
consolidate, KSC must seize the opportunity to streamline the operation. 
It seems that there are so many changes to the procedure that the 
computer programs may need a major rewrite. In the quest for stable 
work flow procedures, a major seismic tremor has been generated that 
will seiid shock waves through the computer systems for some time. 
However, as with ground faults seeking equilibrium, a mo»-e stable fvture 
computer base is the eventual derivative. 

The second approach thaL attempts to pour more resources into 
the computer department so that edifications can be made quicker and 
easier, can certainly reduce ♦■he backlogs. However, a number of 
pracft ;al issues limit a total -.-'lution by using this approach. Buying 
major upgrades to computer system^; is a very time consuming task due 
partly to the government procurement regulations. Increasing the staff 
is sometimes even harder due to the shortage of computer personnel. 
These two constraints prevent sizing the resources to equal the task. 
As Figure 1 shows, the limited resources applied to the requested 
modifications tends to flatten the need date curve into an 


■■"'Z'^^SSP '-. 




'■>t H-T '^ • \. 
















Figure 1 



.v-'.ri.S3|, ..i:. 

implementation curve that closely resemblep the activations of resources 
curve. Almost all of the requested modifications become backlogged 


Where Are the Priorities - Where Should They Be ? 

The computer systems that have been described are Level IV work 
procedures. Information from these data bases feed computerized systems 
at Level III (e.g., Artermis schedules) and Level II (e.g., "Inter- 
Center Problem Reporting and Corrective Action"). The Level II data 
bases are used for program management, the Level III for planning and 
integration, and the Level IV for Implementation. When the computer 
systems are down, the ability to get the job done is impacted at all 
levels. When the requested modifications are backlo£;,ged, the jobs take 
longer to perform. Impact to the computerized procedure directly 
affects the productivity of the work force at each level. The dreaded 
impact to workforce productivity tends to place a priority on 
modifications that benefit the work force rather than the modif icatiois 
that benefit the managers. The reports generated from the data bases 
provide data to the people who do the work. Reports designed to 
identify trends that would be uf*^ '1 in making management decisions are 
not prevalent. Normally, professional and technical people provide 
management with oral and written reports that summarize progress or 
identify problems or issues. The data bases that support the vork force 
could also provide valuable information. Unfortunately, these reports 
in their current state are usually bulky and hard to interpret. Some- 
times the information is scattered across systems and computers and is 
very difficult to integrate. On top of all of tnese problems, they must 
be mailed or hand carried. Often the information is badly dated by the 
time it hits the mail drop. 

The Office Automation Alternativ r 

Office automation may oe the answer to the modification bottle- 
neck and the awkward managem'-.nt reporting system. If managers or their 
executive staffs had access to personal computers equipped with software 
tools to manipulate data, and these tools were networked to the large 
DBMS, then reports could be tailored to the individual manager and 
delivered electronically to the local office printer. As the staff 
becomes more familiar with the information in the DBMS and learns more 
about the power of the tools avail -•*>le through the personal computer, ad 
hoc reports designed by th^ staff can generate timely responses to 
immediate requirements for information. By expanding the hardware and 
software tools, both managers and workers can tap the information to 
suit their need? without impacting one another. 


Two Obstacles That Can Be Eliminated 

In order to be effective, two major problems outside the office 
automation system must be solved. First, the various mainframes that 



fc wik,A-,i,„ -.u :.•« 


host the large DBMS are either currently overloaded or operating 
marginally during periods of peak utilization. If office automation 
demands are to be met, then long range mainframe utilization patterns 
need to be studied and adjusted to accommodate the traffl:. The office 
automation system could provide central hardware that wodd relieve a 
portion of the loads on the mainframes. The second major problem that 
' needs a solution is tha outmoded KSC communication plant. NASA and 

contractor personnel are concentrated in two major areas that are six 
miles apart. The Kennedy Switched Data Network (KSDN) that is currently 
being installed will provide the communications backbone between the 
major buildings and population centers. This system is basically a 
multiplexed twisted pair solution that will maximize the utilization of 
the existing cable plant. It will serve the communications requirements 
until growth pushes the Center toward a fiber optics replacement. Local 
area networks as part of the office automation system would solve some 
of the rigidity of the KSDN's twisted pair solution. The current 45 day 
lead time required to attach end user equipment to a twisted pair cable 
plant "ould be eliminated by providing lo:al area network outlets in 
— each room. The local area networks within major population areas and 

the KSDN between areas would network end users to any destination 

-" The Coals of Office Automation 

f There are a number of committees throughout NASA devoting their 

time toward achieving increased productivity through improved management 

■^ information systems. Figures 2 and 3 identify the NASA Goals and the 

?i NASA-wide information system steering groups. Office autojiation assists 

in the achievement of all of these objectives. 

"I On the local level, KSC must improve the effectiveness of NASA 

personnel in order to meet the increasing demands of the Shuttle multi- 
vehicle processing. Space Station planning, Shuttle/Centaur modifica- 
tions, and various new support requirements. An integrated office 
automation system provides for increased productivity through the 
following gencial objectives: 

o Provides more cimely and integrated information access. 
" o Improvatj communications between workers. 

o Implements a wide range of cost effective office automation 

technologies and applications, 
o Facilitater decision making. 

KSC's Approach to Office Automation 

KSC's approach toward achieving an integrated office automation 

system has been to focus the activity through the Office Automation Task 

Team (OATT) and the ensuing Off:'ce Automation System (OAS) Source 

J Evaluation Board. Since inception in March 1983, the OATT has conducted 

s site visits of installed systems, reviewed the literature, canvassed the 

f KSC community, consolidated the requirements, and defined the specif ica- 

.fc tions. The OAS Source Evaluation Board issued a request for proposals 

\ in January 1985 and a source selection is scheduled for August 1985. 



1^ . »^"r>*k \i--. 






























UJ ui 






Zj CO 


UJ Ol 


< t 

2 CE W 

Euj i^ 

> Q. (_ 

2Z 3 

"^ < s; 

UJ CJ 2 

c; i^ ^ 
^ I o 

< t Q. 
UJ "^ CO 


U M cc 

<l- -J 


—J UJ a. 
Og O 
cc UJ Q 


u. < < 

"J ? !?■- 

Q < ? 


DC Z — 








Ui UJ 



CO 2 


U - 



















UJ o 

> *" 













o ^ 

< UJ 
UJ < 
















UJ ^ 









f- UJ 


CO < 


~- 1- 

z 3 

UJ u 

cc 111 


D H 








2 _ 


CC ^ 
0. UJ 


I ~ 
















f 2 


cc > 
o < 

























o a- 
UJ o 































UJ > 
<J CO 

CO < 


















Ol CO 


U. . 


o o 



















a. <a. 






>J ^ 
















U. r- r- _ 


































QC . 

rr UJ 

CO >- 















> o 






0. Z 

< < 




a. r- 




































Ui X 


CO o 

< < 

















^k^*b«^ *'X,,-- '>f' 




















Wl VI 



<U t.1 

4) X tn 


C T3 0) 

H 4J X 


O 3 iJ 

3 .-O 


•H iH Vj 

tfl M (fl :x 

•r^ a 


iJ O CO 

C O -r^ « 

> -H 


CO C 3 

4) VM £ U 

•H £ 


4J •!-• C* 

H to 

4-1 tn 


C "O 

O >^ 60 

O ^4 


<U Q. CO 

4J U 

CO 4) 


e -H (u 

■H • Vj 

^ • 


4) J= X 

tn ^j ^ o 

a. E tn 


•-I tn 

4J 3 )-i x-i 

Q 4> U 


ex »4 Ul 

V- o o 

< s: 0) 



3 4i 3 in 



•W ^ .H 

T3 01 AJ C 

(4-1 ^ 

<u w 

B O. 

4J 4J CO 

O • CO 

o w 

73 O 

u -o z -^ 

tn 3 

■ri 4) 

C S 0) 


4J M c- 

U-1 U 

CO l-i 

>-i CO C 

j; 4) -O 


U-l U 


Q. O M 

60 4J CO 


u . u 

>, -w flj 

1-1 u 41 




■Q iJ -U 4J 

to CO X 


t-i o- 

t' E 0) 

C 1-1 to c 

U F. 


O T) 

E <U O 

to i-t O 4) 

W T3 


uh CO 

O. J-i 

•rl -H O 

> >^ C 



O « i-l 

tn ^ c 

o y CO 


c s: 

<-! >, r-l 

4) CO 3 60 



(U to CO 

C M g c 

4) .-( tn 

■H T3 


•H a> E -H 

■3 O V- 

iJ C 

a) c- i« 

.H a c 4-1 

1-1 Q. 4) 

rj CO 

T) Q o 

4) O O to 

3 4J 



T3 W C 

1 c c 

u <n 

0) x 

T~t 9i iJ -ri 

>% o a 

o u 

^ cu a) 

3 4J V. -O 

o o 

u^ <U 

4J T3 > 

60 d o M 

C tn 

C iJ 

-H -H 

■H C O • 

4J iH iH 

•H c 

tn i 4J 

tn CL o 4J 

00 tn rH 


(1) 1 to 

4) -O 3 U >H 

to to CO 

u) U 

AJ >, iJ 

J= c to o 



to u c 

t.-, 01 tn o. 

u) a. tn 

00 c 

c c a. 

•H 1 E a (i. 


c « 

■H 0) (T; 

—1 O 10 -O 3 

-0 a T3 

CO <u 

-a M d) 

XJ 4J ^J 3 tn 

1-1 3 

j: S 

v. < t. 

CO 1 IX. -1 

> .C r-l 


C G. 

4J T5 o tj t; 

O 4J tl 

X (U 

O Wi HI 

tn c t- c c 

M 1-1 C 

U X2 


w a. vH to 

a. 3 -H 



























































c c. 




t-i 3 







(J Wi 




•H O 









O 60 

4-1 ^ 

E ^ 



v- c 

to •r^ 

CO V4 



U T^ 

E a 



U Jd 

o c 

60 3 


0) u 

AJ 3 

O iJ 


r-i O 

3 O 

l-l i) 


w 3 


< U 



C- 2 








t— ( 















,■^1 .^Vi^3V\K 

Practical experience with networked office automation systems 
has been obtained through a leased pilot OAS that is networked as well 
as connected to a system of data phones in key management areas. A 
personal computer loan pool has been established to promote the use of 
automated techniques. As with most government and non-government 
organizations, KSC has previously spent its office automation dollars on 
word processors for the office support personnel. Now that communica- 
tions networking for stand alone units is becoming more available, the 
future targets for Increased productivity are the managers and profes- 
sionals who account for 80% of the total office personnel costs. While 
stand alons personal computers can increase the productivity of this 
group to some extent, the timely integrated reports from the large DBMS 
will provide a majOr portion of their decision support system. 

Without listing every office automation technology that KSC 
expects to get, there have been a number of features that have been 
identified as critical to system acceptance by the KSC community. The 
office automation system must have a graphics package suitable for 
generating visual aids of moderate complexity, must have an integrated 
approach to the office systems functions, must be user friendly and 
responsive, must have a powerful electronic mail and filing system, and 
must have a comprehensive data base manager and communications 
cppability. A major goal is to provide the networking functionality to 
the KSC contractors' office automation systems. Adequate training is 
viewed as a major key to user acceptance and system success. 

Office Automation Expectations 

Integrated information serving all levels cf the work force and 
management is KSC's expectation. Planning and reporting are expected to 
shift from "anticipatory" to "on demand." Planning will shift from 
analysis to simulation. Reporting will shift from historical trend 
projections to real time control. Information will become more 
accurate, more detailed, and more available. People at all levels will 
become more productive. 

On the other hand, the management of expectations is a critical 
success factor for office automation. How fast can new technologies be 
absorbed without disrupting the work force? Technology is a moving 
target - there will always be more tomorrow. There is a critical need 
to promote the acceptance of lags between the creation of technology and 
its implementation and between commercial availability of technology and 
meaningful user absorption. The KSC implementation plan seeks to avoid 
disruption, protect investments, secure acceptance, justify costs, 
provide functionality, and prevent obsolescence. Office automation is s 
process rather than a project. The office automation user for the first 
time will have the opportunity to solve the cumbersome manual procedure 
through automated methods. As the work force experiments with the tools 
that are available through office aut^^ation, they, the end user, will 
invent the office of the future thrcgh the natural selection of the 
useful features. 





During her 21 years with the John F. Kennedy Space Center as a 
mathematician/electronics engineer, Georgia H. Brock has accrued 
experience across the spectrum of the information management discipline 
from computer systems analysis to management of large STS ground 
operations data bases. Her current assignment is to manage the project 
planning, engineering, and implementation of the KSC centerwide Office 
Automation System. 





Don F. Simanton and John R. Garman, Johnson Space Center, Houston. Texas 


Information flow is a critical element to intelligent and timely decision-making. At 
NASA's Johnson Space Center the flow of information is being automated through 
the use of a centralized backbone network. The theoretical basis of this network, its 
implications to the horizonal and vertical flow of information, and the technical 
challenges involved in its implementation are the focus of this paper. The importance 
of the use of common tools among programs and some future concerns related to file 
transfer, graphics transfer, and merging of voice and data are also discussed. 


One of the cornerstones of quality and productivity is good informed decisions. The 
National Aeronautics and Space Administration is a civilian agency of the U. S. 
Government which has been engaged for over 25 years in research and development 
for air and space flight. The I'ohnson Space Center ( JSC) in Houston, Texas, as one of 
NASA's nine field centers, has become the leading center in project and program 
management and space flight operations. Data systems have supported the decision- 
making processes at the Center since its inception and have continued as an integral 
part of the Center's activities for over 20 years. The last iO years have brought rapid 
growth in computer technologies which in turn has enabled ever-increasing 

sophistication in the use of computers in the operational and management decision- '-. 

making process. Capitalizing on these technologies is no* only a challenge for JSC in i 

striving for increasing quality and productivity in its produ'^ts and activities, but is a i 

major necessity due to the increased scope of JSC responsibJities. JSC is the lead 
NASA center for both the new Space Station Program and the on-going National 
Space Transportation System or "Space Shuttle" program. 

In order to ensure an integratea approach and to provide direction to the growth of 
data processing at JSC, a strategic planning committee for automatic data processing 
( ADP) was established in 19S4. Many of the concepts included in this paper are a result 

of the work of that committee which was presented in a formal report in the Spring of ' -^ 

1985. Y 


Information flow requires three basic elements; (1) a need to transfer information, ■ 

(2) a physical connection, and (3) understandability. For example, consider two ! 

persons speaking to each other over a telephone. First, they must wish to have a 
conversation. Second, they must have a connection (i.e., the phone must be in order, 
the line must not be busy, and both persons must be present). Finally, they must be 
able to understand each other, as in speaking the same language and having a similar 
knowledge about the topic being discussed. 

I'ransferring information in a data processing environment utilizes a similar 
process that also involves all three attributes. This paper deals primarily with the 
first two steps in the transfer: the need and the physical connection. More specifically 
it deals with the theoretical basis of the need and the application of that theory at JSC 
together with the theory of the management of a network and the way management 
has been accomplished at JSC. The tasks left to be accomplished and the subject of 
understandability will be touched on at the end of this paper. 



■-'if-v^ -, ♦' 



using an adaptation of Robert Anthony's analytic model cf an 


In an ideal organization, information flows unimpeded both horizontally (laterally) 
and vertically through the management structure. This flow can be best expressed by 
using an adaptation of Robert Anthony's analytic modej of an organization [1] shown 
in figure 1. This model includes three levels of management structure: the strategic 
level, the tactical level, and the operations level. The information needs of each of 
these levels differs considerably. For example, the operational level is concerned 
with detailed information, the tactical level with summary information, and the 
strategic level with trends and projections. 






•" USER 

Figure 1.- Organization information model. 


Traditionally the data processing industry has concerned itself v/ith the 
information needs of the operational, or lowest, level. This is t^e level of "automation" 
where machines are used to enhance, or in some cases to take ever, processes 
performed by people. Data processing has done well in this area, particularly in 
taking over mundane tasks involving data manipulation and in providing higher 
levels of accuracy and quickei responses than are possible in a totally manned 
environment. People using data processed at this level are familiar with the 
processes and the data processing required for their specialized functions and are 
normally characterized as sophisticated users of data processing. 

At the tactical level the data processing industry is not as well established. Typical 
products today BlTp, standard summary charts produced by overnight "batch" 
(non interactive processing) jobs. This type of reporting offers little flexibility in 
information available to users. However, this situation is being addressed by a new 
generation of application development tools, termed fourth generation languages, 
which have recently evolved into viable commercial products. These languages 
permit information queries on a relational basis thereby allowing questions to be 
unstructured and interactive as opposed to the rigidity imposed by the "hard coded" 
batch report approach. It is this unstructured aspect of the languages that has given 


- -.S^Ji^ -v 

Hh .\v,v:i;%vt-, 


their, the name "decision support" languages. For managers can now request more 
than data. They can request decision information, i.e.. data relationships or 
information from data processing reports which meets specified criteria. These 
Hnguages are also more "user friendly" than those found at the operational level. 
This characteristic enables less sophisticated users to be effective "programmers of 
applications." It transfers some of the workload or "cost of ownership" of application 
software from providers to users, which in turn creates higher productivity and 

Until recently, the data processing industry has had little more to offer above the 
tactical level. Standardized summary charts were the extent of the available 
technology for strategic planners. Trend data could be produced but based only on 
previously defined criteria. Fourth generation languages and specialized packages 
which produce trend data and allow "what if" questions are the beginnings of what 
promises to be a set of powerful tools for the strategic level. 

The model created above gives two driving forces for providing communication 
linkages wchin a given organization. One linkage is horizontal and primarily 
supports automation in the operational arena. The other linkage is an emergent one 
and provides a vertical flow of data through the organization. There is also third 
driving force, not clearly shown in this model. This force is one of providing common 
tools for common uses. 

The Johnson Space Center can be viewed in several ways. It may be viewed from a 
project sense wherein there are two main projects and host of smaller ones (the main 
projects being the Shuttle Transportation System and the Space Station). It may also 
be viewed as an organization consistmg of project offices, an administrative 
directorate, an engineering directorate, and an operations directorate. A third visw 
might be to consider it as a collection of engineering systems, operations systems, 
and information systems. Figure 2 shows how these systems relate in an integrated 

A tendency in any organization is to organize vertically, such that all functions 
needed by an organization are integral to that organization. For example, ail 
organizational elements have need for a budget system. Thus, vertical organization 
at JSC provides that for each organizational element the budget system be contained 
within that element. This leads to separate and distinct budget systems for the project 
offices, the administrative directorate, the engineeering directorate, and the 
operations diri'ctorate. To make matters worse, these separate budget systems might 
also be duplicated by the program (Shuttle and,' or Space Station). Hence, this 
structure is referred to as a "4 X 3 structure" in its tendency to create 4X3 sets of 
systems for each required function across different systems. 

In the center of figure 2 [3] appears a logo for th^ JSC Center Information Netv/ork 
(CIN). The logo implies a concept whereby the 4X3 problem is avoided by 
interconnecting the data processing systems to allov,; the use of a shared supporting 
system across organizations and projects. This is the "common tools for con.mon 
purposes" concept, which is the third driving force for centralized communications 
linkages. Conceptually it is the inverse of the more traditional federated or 
"departmental" systems approach. Instead of bringing complete sets of unique 
applications and information to each user group, all user groups are provided access 
to common applications. While the information structure remains highly federated 
("you can't see my department's budget until I'm ready to submit it," i.e., move it up 
the triangle), the interfaces and bridges required for both lateral access and vertical 
integration are significantly reduced. 



♦J ^i1^ 


or;g;:^al paoZ u> 
of poor quality 

-4 om^ATiONS srtrrMs i > ' , i4 





Figure 2.- "Four by three" structure. 


Over the last year the forces described above have created at JSC a great emphasis 
to build a multinetworked series of interconnections enabling any element of the 
Center's data processing equipment, particularly users' workstations, to 
communicate with any other element of data processingequipment. particularly that 
which hosts common applications. Because of the universal nature of the three 
drivers listed above, it is anticipated that this networking will be extended to ail of 
NASA's centers within t^e nt ■♦ two years. 

The Theoretical Basis of Interconnectability 

JSC uses six principal ways to network and manage data processing equipment [3]. 
Thesj are: 

• Connecting equipment of like architecture via front end coinmuni(~Ations 
processor backbone linkages when a single organization has con.plete control 
(centralized management) of all part:> of the network, including attached systems 

• Connecting equipment of like architecture via front end communications 
processor backbone linkixges with centralized management of the enviror ment 
limited to control of the network only. 

• Connecting equipment of different architectures via hardware and/or software 

• Connecting equipment of like architecture via separate front end backbone 
comniunicaticns processors such that there is severability between networks. 

• Connecting a terminal to more than one network via multiple interfaces at the 

• Connecting networks by using a limited bridge. 




Each of these methods is illustrated in figures 3 through 7 (the latter two methods 
are both shown in figure 7). In these diagrams the network area of responsibility is 
depicted by the rectangular outline in the center. The usercomm.inity is shown on the 
left, and the host processors (applications and information) arc shown on the right 
The fixed components of the network are the controllers (to which the terminals 
attach), the backbone wiring, and the front end processors. The variable components 
of responsibility art the bridges and the operating system software 

In figurs 3, called "Class A," the provider of network services is also the "owner" o.*^ 
the operating systeni within the host processor. This arrangement consolidates 
complete control (and service) over the management of the network.and the user need 
only be concerned with the terminal and the application running on the host. 

The "Class B" situation in figure 4 differs from that of Class A in that the 
"ownership" of the operating system is severed from the consolidated 
management control of the network This situation accordingly requires a degree of 
coordination and interface as the network managers no longer have complete control 
of nil of the components necesssary to make the network operate. The network- 
related systems software within the operating system must be considered the 
property of the network manager, who must retain approval authority for chan(;es to 
that software. 

Both Class A and Class B have dealt with two like (architecturally compatible) 
processors "Class C" in figure 5 presents two computers of dissiniiliar architecture 
In this case a protocol converter is needed to "bridge" two architectuids. This 
bridge may be "one-way" whereby one network accesses the processor of he other 
network but not vice versa, or it may be "two-way" whereby terminals on either 
network may access cither network s processors. 

The network shown as "Class D" in figure 6 shows a different method of 
interconnecting two networks. In this diagram a given host has two xront end 
processors, one of which is connected to two separate networks. The networks shown 
are of like (compatible) architectures but could easily involve bridges to dissimilar 
architectures. The advantage of this arrangement is that the secondary network 
connection may be sever'.'d by "downing" the front end processor. An example of this 
arrangement is its use at JSC to ensure that there is no outside access to Shuttle 
mission si'pporting >. ^cessors during a mission. During nonmission periods access 
IS enabled in order to bring in development and maintenance terminals. 

"Class E" in figure 7. for one example, is characterized by two networks connected 
in such a way that each network is not fully aware of the other. In this example in the 
first method a terminal has connectivity to both networks by the use of two sets of 
interface cards The second method shown portrays a limited "file transfer only" 
bridge between the networks. These are termed method 1 and method 2. respectively. 

Connectivity Issues 

The bridging of multiple vendor networks raises questions of vendor cooperati^.:!. 
This issue is one that must be addressed both from a software products point of view 
(some bridge implementations work only if certain products are installed on the other 
vendor's machine) and from a maintenance and problem-troubleshooting 
perspective. Obviously, cooperation is required during installation and also 
whenever a problem occurs which does not clearly belong to a given network 

When government and contractors share a common network theie exists a need to 
provide strict access control both to application programs and to processors which 
may be on the network but which need not be accessed in the course of doing business. 
Such control may be achieved with the type of connectivity method chosen (e.g.. Class 
E IS much more restrictive than Class 3). by tie u^^ of user IDs which have limited 
processor access, by the use of application passwords, or by a combination of all of 

[ these 











SHsHi}®®— -Q- 
B®«aK5] — -Q- 


SHiHiHfHSl — kZh 


B&SiSHi}— ■ 


Figure 3.- Class A connection to CIN. 


(sHSSSHsl — -D 
BSHSSisj— ■€> 
Is)®®®®— O 




B«®S®— -D 

BSttKiKS — -Q 

SHsHSSa— O- 

SSilHi}®— O 

SHsH&S® — -Lr 

SHI}®®® — |-D 

®®S®®— -O 


®®®®a— -D 

B®®®®— -D- 

Figure 4.- Class B connection to CIN. 


ffl®®®®— -D 
®{a®®®— HI} 

B®®®® — -D 
B®®®®— -D 


BBS®® D 

SB®®® D 

B®®®® D 

Figure 5.- Class C connection to CIN. 

Figure 6.- Class D connection to CIN. 



: ■• ''**;^' '••• '' 



ffiSSSffl— fo 

SHsHS®®— Q 

gJHsHsKiHa — -Q 

SyfiHsHlHsl— Q- 
SHsHsHsH*)— 10 


Bsssa — D 

Figure 7.- Class E connection to CIN. 

Security is always an issue with extensive networks, especially those v/hich have 
intP'-faces to public telephone networks (and most large networks have at leasl one 
such linkage). The convenience that access to a public telephone network provides is 
of great value. Infrequent users can dial in via telephone from nearly anywhere and 
conduct business. The problem is that so can any "hacker." Protection measures 
commonly utilized include callback requirements, the limited distribution of access 
numbers, network passwords, and application passwords. Nevertheless, the network 
with an interface to a public telephone network has wide exposure to a misuse of 

Connecting special-purpose local area networks (LAN's) is an emergent 
technology. In these cases the interface to the LAN appears as a terminal to the 
backbone network. Yet the LAN itself may comprise a number of terminals networked 
together. If multiple sessions can occur simultaneously across the interface, the 
topology of the two networks appears similar to that shown as Class C except that no 
processor is involved on one side of the bridge. 

An extension of the LAN concept is that of connection to a wide area network 
(WAN), a network that connects multiple networks together over very high-speed 
linkages. The public telephone system is an example of a very large WAN. NASA is 
currently interconnecting all of its major field centers over a WAN named the 
Program Support Communications Network, This network uses the X.25 packet 
switching onventions over a network of both leased and dedicated T-1 carriers. 

Building the Center Information Network 

Examination and analysis of the many separate networks existing at JSC in early 
1984 led managemsnt to conclude that the best way to integrate the networks was to 
interconnect and build on the existing IBM networks and to standardize on IBM's 
system network architecture (SNA) as the foundation for expansion and integrition 
with other networks. The choice was made based on three factors. P'irst, already in 
place was a substantial set of SNA networks at JSC. Second, most major vendors have 




'*kiM Ol. r"-*, -^ vW" 

'^ . 

*•' or are buiding bridges to SNA. Third. SNA corresponds closely, albeit arguably, to the 

\ open systems interconnection (OSI) m- ■n. The OSI seven-layer network model is a 

^f reality toward whicn all vendors are wo ig, with completion expected in the 1990's. 

,s. The setting of the SNA as a standard having been accomplished, the current 

» backbone network integrates multiple vendor products (e.g.. IBM, Sperry. DEC. 

Xerox) covering most of the Center. Each of the six theoretical types of networks are 
i currently either in use or planned for use at JSC. Where possible the network paths are 

f two-way: however, currrent industry offerings limit some of the paths to a one-way 

f terminal connection. This limitation is expected to be lifted m the next year for most 

* vendors. A policy of "buy if. don't build it" has also been established for network 

growth. Although wails have occurred for needed technologies in some areas, the 
course of events during the past two years has proven the wisdom of the policy. In 
each case the required capabilities have appeared on the market prior to the time JSC 
could have developed an equivalent function "in-house." The product has also been 
far less expensive than would have been the case for in-house development. 


The tasks facing JSC are still enormous and pose a challenging environment. 
Networking is simply the first step which, furthermore, has yet to be completed. 

An immediate goal is establishing true two-way connections. Transferring data 
files is still more constrained than is desirable and thus is not yet a widespread 
capability. As of the writing of this paper there is no standard for document 
interchange (unless standard ASCII is considered) Re visable form documents 
created in one brand of word processor may be electronically transmitted from one 
organizational element to another but they are illegible upon receipt, almost as if a 
speaker were conversing in Greek but the listener knew only English. Substantial 
need exists in this area for a dictated de facto standard supportable by many vendors. 

Graphics data transfer is in even poorer condition as exchange capability from 
vendor to vendor is almost nonexistent. Yet graphics data is destined to play an 
important role in the Space Station, hence it is imperative thr>' the various NASA 
centers be able to exchange graphics data if NASA is to have a well-coordinated 
Space Station Program. In the Shuttle Program, paper has been used as the 
transmission media for graphics data. In this era of increasing responsibility and 
decreasing budgets, however, the productivity and quality gains enjoyed through the 
use of data systems for engineering and technical management are becoming 
mandatory elements of major programs planning and costing. Finally, voice and data 
in combination is just over the horizon, according to the major data processing 
vendors. Yet no plans are yet extant to provide vendors' standards for voice and data 
in combination. 

To accomplish these tasks, for most of which solutions remain to be found, JSC 
management has chosen the tactic of implementing a strategy railier than a plan. 
Such an approach may seem strange to an engineer in that a strategy implies 
certainty about an objective but uncertainty as to how to achieve it, A detailed plan 
does not in fact exist. Rather, prototype efforts will take place, successes will be built 
upon, and failures will be discarded. 

Figure 8 depicts this concept of a strategy-driven approach to building ^nd 
integrating data processing over a common network. An excellent description of this 
process may be found, incidentally, in Bernard Boar's Application Prototyping [2j 


In order to provide management with the information required for intelligent and 
timely decision-making, JSC ADP strategic planning has recommended the use of a 




■ - ""^^^t -.■ 








AOPBG 10?^ iS 85' 


Figure 8.- Strategy-driven approach. 

common network strategy. This strategy has been adopted and a theoretical plan for 
network management has been developed. A common architecture has been selected 
to ensure that connectivity exists among all organizational elements and to provide 
the basis for transferring data and information to all levels of management. This 
universal access to information combined with common tools will create new 
capabilities of information access to the tactical and strategic levels of management 
and will further promote the flow of information at the operational level of 

The products currently supporting implementation ars "state of the art" vendor 
products that engage only the initial aspects of the networking problem, those 
dealing with connertivity. Hence the network engineering task remains a 
challenging and path-finding exercise. Yet the network is an iceberg with but its tip 
showing. There remain the problems of understandability, of transferring graphics 
data, and of the integrated transfer of voice and data. As the networking problem 
evolves, corresponding management challenges will emerge from which new 
strategies will evolve for dealing with the problems of centralized communications 


1. Anthony, Robert N,. Planning and Control Systems: A Framework for Analysis, 
Harvard University Press, 1985. 

2. Boar, Bernard, Application Prototyping, John Wiley & Sons, New York, N.Y., 1984. 

3. Dunseith, L., and Garman. J., "Report of the JSC ADP Strategic Planning 
Committee," Internal Presentation. NASA Johnson Space Center, April 1984. 



_ > L - ' 


Don F. Simanton is the Chief of the Systems Planning Branch of the Data 
Processing Systems Division at NASA's Johnson Space Center. He has been 
responsible for the engineering of the Center Information Network at JSC. He also 
was a member of the JSC ADP Strategic Planning Committee. 

John R. Garman is the Chief of the Data Processing Systems Division at the 
Johnson Space Center. He was a member of the JSC ADP Strategic Planning 
Committee and was a principal in the implementation of Shuttle avionics flight 
software and the tools used in its development. 

> 211 



*i *W.V.'' ~^ NU l"^ ' 

« '- 1 N...1. 



ih»ti**it.v8^i5>»v;.v '^^^ifts 


Richard L. West, McDonnell Douglas - Houston 


This paper describes a performance measurement system being 
developed by McDonnell Douglas Technical Services Company-Houston. 
Based on the Family of Measures concept proposed by the American Pro- 
ductivity Center, the measurement system provides both a means of moni- 
toring performance and a resource to support management decision 
making. The process of performance indicator development is discussed 
and typical indicators are described. 

Th(! paper concludes with a summary of some of the lessons 
learned in applying productivity measurements to engineering services 
tasks and in automating data collection, evaluation and interpretation. 


McDcnnell Douglas Technical Services Company-Houston (MDTSCO-H) 
built its fo'indations at JSC on a small task in 1965 to convert the 
Gemini Missio.i Simulator to an Apollo Procedures Trainer. Since that 
time, MDTSCO-H has provided engineering and operations support on 
Apollo, Skylab, the Apollo Soyu: Test Project and the Space Shuttle 
Program. Over 1000 personnel are now a part of the Houston Operations, 
providing support to JSC under the Space Transportation System Engi- 
neering and Operationa Support contract and supporting McDonnell 
Douglas Space Station Phase B studies,. Ninety percent of the current 
employees are engineering and technical people performing non- routine 
tasks for the nation's apace programs. 

In 1982 the company proposed to NASA a streamlining program to 
reduce STS costs and improve mission effectiveness. In October 1982, 
that initiative was formalized under the STS Engineering and Opera- 
tions Support Contract [?]. In early 1983, McDonnell Douglas corporate 
self-renewal initiatives were imp]emented and formed the basis for the 
current Quality anci Productivity Improvement Program at MDTSCO-H [3]. 
One element of that Quality and Productivity Improvement Program is che 
Performance Measurement System discussed in this paper. 



■ -.'*r^- 



The MDTSCO-H PMS is based on a definition of performance im- 
provement that includes improvement in both quality and productivity. 
Development of a PMS as a key part of the Quality and Productivity 
Improvement Program was motivated by a number of factors: 

- An effective PMS provides a highly visible indication of "how 
goes it" and a resource for management decision making. 

- It allows management to diagnose the past, control current 
performance, and plan for the future. 

- It allows management to evaluate the effectiveness of per- 
formance Irprovement initiatives by correlating changes in 
performance indicators with implementation of those 

- Finally, and perhaps most important, it provides a means for 
keeping everyone in the organization informed and involved in 
the continuous improvement of performance. 

With such a list of payoffs for a PMS, the motivation for its 
development and implementation was clear. But, although the motivation 
was clear, the approach to development and implementation in an engi- 
neering services environment was not. 


The initial effort to develop a PMS at MDTSCO-H was undertaken 
by the Performance and Productivity Panel of the Quality and Productiv- 
ity ■improvement Council (QPIC). The council is comprised of the top 
three management levels at MDTSCO-H and is the focal point for quality 
and productivity initiatives [3]. 

In reviewing available literature on quality and productivity 
measurement, the Performance and Productivity Panel found extensive 
references related to a production manufacturing environment. In such 
an environment productivity was easily defined as output divided by 
input - widgets per dollar or widgets per hour. Quality was defined in 
terms of scrap - defective widgets per thousand. 

Early efforts to apply such definitions to an engineering serv- 
'ces environment proved difficult. How does one measure productivity 
in an environment where most products are unique; where the ascent 
flight design for one mission may be significantly more complex than 
one for a superficially similar mission? How does one quantitatively 
define scrap in such an environment? After several false start . the 
panel was introduced, through an American Productivity Center seminar, 
to the family of measures concept proposed by Carl G. Thor [^4]. 




The APC Family of Measures Concept 

Mr. Thor pointed out that "it is necessary, much inore frequently 
than in a factory, to use a collection of measures rather than a single 
measure for a particular (professional) department." The reference 
provided an example of how the family of measures concept could be 
applied to an engineering department. This family of measures concept 
was adopted by the Performance and Productivity Panel. The reference 
further suggested a participative approach to the development of per- 
formance measurements. That approach became the nucleus for the 
panel's PMS Development Plan. 

PMS Development Plan 

The development plan outlined an approach for development of a 
hierarchical set of performance indicators beginning with definition of 
top-level (division-level) indicators by the QPIC. The top-level indi- 
cators provided a context for definition of related lower-level ones. 
These lowei — level indicators were augmented by others unique to the 
particular organizational element for which they were developed. 

Development of Top-level Indicators 

The Performance Measurement System was intended to pro/ide both 
hard and soft indicators of performance. Hard indicators included the 
quality, productivity and timeliness of products, and quantitative 
ii.odsures of fiscal performance. The soft indicators focused on per- 
formance on corporate initiatives in areas such as participative man- 

-f agement, employee development and recognition, and other factors of 

r quality-of-life in the workplace. 


■e Corporate initiatives defined the major areas of performance to 

1 be measured. The QPIC began detailed development of candidate indica- 

I tors for those area3 using the Nominal Group Technique [4]. After an 

initial round-robin brainstorming and "light" editing, the resulting 
( list of indicators was further edited, organized into categories by the 

' director and published for further, detailed review by the full OPIC. 

The initial set of division-level indicators was baselined and imple- 
;. .nentation begun only after thorough discussion, redefinition, aaditions 

t and deletions. 

i Resulting Top-level Indicators 

I The Division-level indicators (figure 1) consist of 31 indica- 

':■■ tors in seven categories: Business Development Effectiveness, Fiscal 

[ Health, Staffing Adequacy, Contractual Performance, Organizational 

Climate, Organizational Productivity/Quality of Work, and Technology 
•f Posture. A final indicator, Self-Renewal and Continuous Improvement, 

!■ is calcul:.ted as a weighted average of scores in the seven categories 

and provides an assessment of overall performance at the aivision 



7i . .V 







- Vmv Buain««s Funds Expanded 
p«r Dollar Salas 

- Parccnt of Narkat Capturad 

- Addition* to BacUog 

- Salaa aa Parcant of Salaa 
at Stalca 


- Salaa 

- Eaminga 

- Oiract Labor Rata 

- Ovarhaad Rata 

- Coat par Manyaar Equivalent 

- Ganaral and Adainiatrativr 


- Parcant Staffed 

- Years Since Degree 

- Percent of Staff with 
Advanced Oegreea 

- Grade Level Distribution 


- Percent Authorized Houra 

- Perfot-aance Evaluationa 
(CPAF Contracta) 

- Fee Deviation froa Noainal 
(Fixed Price Contracts) 

- Cverrun/Underrun of Contract 


- Attrition Rata 

- Absantaaisa 

- Problaa Solving Teas 

- Foraal Awards and Recognition 

- Parcant of Tiaa Spent in 

- Nor)cplac« Visits and Bova 

- Parcant of Workforce Trained 
in Juran-Deaing Techniques 

- Parcant Supervisory Openirtgs 
Pilled froa within 


- Value of Streaalining Savinijs 

- Emiloyae Suggestion Awards 

- Resources Required for 
Plight Preparation 

- Accepted Juran Project 


- Percent of Contracted Work 
Involving Advanced Technology 


- Weighted Average laproveaent 
in Indicatora 

The Organizational Climate category illustrates some of the soft 
indicators mentioned earlier. For example, the "Workplace Visits" 
indicator (figure 2) represents a measure of the level of "management 
by wandering around" (MBWA) as discussed by Peters and Austin [1]. 
While not an end In itself, the Indicator Is expected to allow assess- 
ment of the effectiveness of MBWA In the MDTSCO-H organization when 
correlated with other measures of organizational climate. 

Hierarchical Development of Lower-Level Indicators 

With division-level indicators defined, the next level of the 
organization - Engineering and specific Projects - began definition of 
their PMS indicators. Some were defined automatically as a flowdown 
from the division-level indicators (figure 3). Additional indicators, 
each specifically tailored to a particular organizational element, were 
defined using the same nominal group technlqua described earlier for 
division-level indicators. For example, indicators unique to the Soft- 
ware E.igineering Department Include software development productivity 
and conformance to software standards. 






K\l PLO '-JTS 


< 1 , ■ ■ 

M J ^ , A ^ 




























•••■'rfvjfj ' 

.v*r'c%v^^ '-:'-^ "^ ' ' f^t 


Since definition of the division- level indicators was completed 
only last Fail, and the lower-level indicators even more recently, 
there are a number of lessons still to be learned. Experience to date 
with the MDTSCO-H PMS does allow some generalizations to be made, how- 
ever, about what seemed to work effectively: 

- Start at the top. This not only demonstrates top-management 
commitment but forces careful top-level thinking before in- 
volving a large number of people. In addition it provides 
the required top-level definition of indicators which lower- 
le'-el Indicators must support through the flow-down process. 

- Once the Initial definition is complete, bpgin implementation 
with a few selected Indicators rather than tackling the whole 
list at once. Experiment. Try it out. Then use what was 
learned to expand Implementation to the complete list. 

- Don't restrict Indicators to those that can be plotted as 
control charts. Control charts may apply for most Indicators 
in a production manufacturing environment; in a non-routine 
knowleJge-based engineering environment, statistical bounda- 
ries for control chart limits may not be available. 

- Don't feel constrained to cast Indicators in concrete. Re- 
view them periodically. Delete and add indicators to Insure 
that the PMS really measures those factors that are key to 
the current and future success of the organization. 


Attempts to automate the generation of PMS indicators have in- 
cluded several CSPEO-related activities for DoD applications and com- 
mercial software packages. Most of the division-level indicators are 
currently plott3d using graphics programs on personal Computers. Some 
of the indicators use a standalone program; others, depending on the 
complexity of calculations involved, are generated with spreadsheet 
programs which include an integrated graphics capability. Hierarchical 
measures are well suited to computer-generated spreadsheets and a num- 
oer of such products are available which allow efficient consolidation 
of lower-level indicators to generate higher-level ones. 

Generally speaking, however, the recommended approach is to 
concentrate initially on definition of meaningful indicators and not 
commit too early to automation of an immature set of measures. 

A futur-e goal is to automate as much of tho PMS process as eco- 
nomically feasible. Fxpert systems may also prove to be of assistance 
in decision making using the indicators. For the present, however, the 
MDTSCO-H PMS is primarily a hands-on, manual system, under Internal 

evdluat ion. 





The MDTSCO-H Performance Measurement System represents an effort 
to measure performance for an engineering services organization. In 
addition to measuring the usual hard parameters such as sales, return 
on Investment, etc., the system is designed to provide an indication of 
performance in other areas of corporate initiatives. 

The family of measures concept and use of the nominal group 
technique were found to be effective means for defining performance 
indicators at all levels of the organ i ■nation. The current incicators 
represent a beginning. They, like the organization whose performance 
they meaaure, will evolve as MDTSCO-H presses forward on the path of 
continuous improvement. 


[1] Peters, Thomas J. and Austin, Nancy K. , Random House, New York, 
1985, pp. 378-392. 

[2] Petersburg, Ronald K, , "Streamlining: Reducing Costs and Increas- 
ing STS Operations Effectiveness," Proceedln^a , R&D Productivity 
Conference, Lyndon B. Johnson Space Center, 1985. 

[31 Ruda, Roger R., "Implementing Quality/Productivity Improveiront 

Initiatives in an Engineering Services Environips .t," Proceedings , 
R&D Productivity Conference, Lyndon B. Johnson Space Center, 1985. 

[4] Stephenson, Blair Y. and Franklin, Stephen G., "Bettar Decision- 
Making for a 'Real World' Environment," Administrative Management , 
July, 1981, pp. 24-38. 

[5] Thor, Carl G. , "Measuring the Productivity of Technica' Re- 
sources," Knowledge Worker Measurement , American PrcJuctivity 
Center, ^9W. 


Richard West is a Principal Staff Engineer with McDonnell 
Douglas Technical Services Company-Houston. He has 26 years of experi- 
ence In the aerospace industry. In his current assignment as Assistant 
to the Director, he is responsible for coordinating the development and 
implementation of Division-Level indicators for the MDTSCO-H Perform- 
ance Measurement System. 







Charles W. McKay 

Rodney L. Bowa 

University of Houston Clear Lake 


The 9pace station data management system lnv^,lves networks of 
computing resources that must work cooperatively and reliably over an 
Indefinite life span. This program requires a long schedule of modular 
growth and an even longer period of maintenance and operation. The de- 
velopment and operation of space station computing resources will in- 
volve a spectrum of systems and software life cycle activities distrib- 
uted across a variety of hosts, an Integration, verlflcationj and vali- 
dation host with test bed, and distributed targets. This paper identi- 
fies the requit.^ment for the early establlshnent and use of an appropri- 
ate Computer Systems and Software Engineering Support Environment. This 
environment will support the Research and Development Productivity chal- 
lenges presented b^ the space station computing system. 


The Space Station Program is different from past NASA projects 
in that It will provide the capability to support a permanent manned 
presence in space. The computing resources will provide an end-to-end 
Information system that will evolve over approximately 20 years and re- 
main operational for an indefinite period. At maturity, the Space Sta- 
tion Program will Involve networks of computing resources located on the 
ground, in low earth orbit. In higher orbits, and hopefully on a perma- 
nent manned station on the moon. There will be vehicles operating be- 
tween the earth, multiple space stations, free flying platforms, moon 
and deep space. Such a program is likely to require the largest number 
of processors ever embedded in an Integrated end-to-end system. These 
processors must work cooperatively and reliably to maintain the system's 
Integrity and quality of services in spite of their physical separation 
in space. More than ever before, the key to building and operating an 
adaptable, reliable system Is believed to be the early establishment and 
use of an appropriate Computer Systems and Software Engineering Support 
Environment. This environment will support the technical response to 
the research and development challenges of the NASA space station com- 
puting system. 

♦Ada is the trademark of the US Government, Ada Joint Programming Office 


:'*r3f^' ■-.-' 


5J?-Pt, S^'-. .\. - " K^J 


Parallel Research 

The space station data management system can be divided Into 
three functional areas: 

1. Network Communication Services (NCS) 

2. Network Information Services (NIS) 

3. Network Applications Services (NAS) 

This functional division will maximize the opportunities for parallel 
research In related areas of conceim. This approach is illustratr^d in 
Figure 1 which shows three types of Local Area Networks (LAN's) expected 
to be In day-to-day use by year 2000. 

The scenario of Figure 1 exhibits Space Station #1, Ground Sup- 
port Station #27, and Manned Orbital Transfer Vehicle #3. Each local 
area network is shown to be functionally divided into clusters. Each 
cluster is made up of three functional components: NCS, NIS, and NAS. 
The lowest layers of the NCS component can be connected to a local area 
network media (eg, a fiber optic ring) to a gateway to another LAN or to 
a transmitter/ receiver to remotely located LAN's (eg, a radio fre- 
quency link or a laser beam link). These lowest level NCS connections 
form stable interface points, if they conform to emerging International 
Standards Organization's Open Systems Interconnection (OSI) model. 

The mort challenging (and Important) stable interface point for 
the purpose of facilitating the farallel research activities is the on^: 
at the Junction of the three cluster coiiponents. A definition cf the 
software and hardware protocols, that govern exchanges among the three 
components should allow researchers to combine the best of their work 
for study. This assumes that each group of researchers applies modem 
principles of computer systems and software engineering with regard to: 
layered design, management of abstract objects, and provision of fault 
tolerance within an environment containing both real time and data 
driven applications. 

Proposed 1990 En vironment 

The best baseline for a 1990 software engineering support envi- 
ronment is a mature Minimal Ada Progamming Support Environment (MAPSE). 
However this minimal toolset is insufficient by itself. By 1990 there 
should be a commonly accepted model and vocabulary for the project data 
base and for enforcing the configuration management policies. This com- 
mon model and vocabulary make possible the development of: a reuseable 
component library applicable to every p.iase of the life cycle; highly 
automated technical tools which r<;enforce methodologies appropriate for 
the various life cycle phsses; and highly automated management tools 
which complement the technical tools. 

As shown In Figure 2, this proposed 1990 environment should sup- 
port (:he continual evolution of the system engineering requirements. 
When enough of rhe system engineering requirements are sufficiently 
understood, work can proceed In parallel to begin to determine th3 soft- 
ware engineering requirements. Continuing thla highly iterative feed- 


. -.,#^^_. ,- 



back process, the work can progress to determine the hardw^^re engineer- 
ing requirements. Throughou*- all three activities there iu a need to 
manage people and logistics. As summarized at the bottom of the Figure, 
this is a highly Iteratlsre process of considering syscems, software and 
hardware where all information for all phases of the life cycle are In- 
put Into a CAIS (Common APSE Interface Set) conforming project database 
with well understood configuration change management. The engineering 
process Is aided by a highly automated set of advance technical tools. 
In turn the appropriate use of the advanced technical tools are reen- 
forced by a highly automated set of complementary management tools. 
This envronment will also support the changing maintenance requirements 
over tne entire life cycle of the data management system. 

Wo rking Definition 

The upper half of Figure 3 proposes a working definition for the 
software engineering process Just described (adapted from Ross, Good- 
enough, Irvine, 1975 [4]). The key to this definition Is the recogni- 
tion of the five goals of uoftware engineering (le, not just "correct"). 
It Is most unfortunate that the rush to produce code which can be demon- 
strated to pass acceptance tests often produces software at costs which 
are typically less than 20Z of the life cycle costs of the software. 
Over 80Z of today's software costs are Incurred In maintenance and 
operation attempts to modify the software to meet changing requirements, 
to repair the softwa-e, and to tune the software. The difficulties 
associated with modifying software In a safe and reliable manner that Is 
correct, efficient and understandable were among the principal concerns 
of the designers of the MAPSE and CAIS. 

The bottom of Figure ? shows the traditional view of the major 
phases of the software life cycle depicted In the textbooks. However 
the top half of Figure 4 Is believed to be a far more realistic view of 
the highly Interactive life cycle phases for projects such as space 
station. As shown, the work commences with the analysis of systems 
engineering requirements. At some point the system requirements are 
sufficiently understood to enable work to begin In parallel on the 
software engineering requirements. This highly Iterative work Is fed 
Into the project data base possibly producing changes In the preceding 
work. These effects ripple through the other phases of the work. When 
the software engineering requirements are sufflcently understood a third 
parallel activity (the analysis of the hardware requirements) can also 

It is Important to realize that the effects of these iterative 
Interactions can be stimulated from any point along a continuum of re- 
quirements analysis activities. At one end of the continuum we deal 
with "illuminated" Issues (things we understand very well). Typically 
we analyze and represent such requirements In a highly procedure driven 
fashion. This gives rise to the 'myth' of proceeding In distinct steps 
from one phase of the life cycle to the next phase In a stairstep or 
waterfall sequence of activities. Unfortunately these are the easy 
Issues to deal with. At the other end of the continuum are the "dark" 
issues. Typically the analysis- and representation of such requirements 
cannot begin In the same procedure driven fashion as the activities at 






the Illuminated end. These nonprocedural activities are more likely to 
be driven by Insights gained from Intuition, Inference and Investiga- 
tion. Rapid prototyping, simulating, and artificial Intelligence tech- 
niques can be very useful In gaining these Insights. 


Minimal Toolset 

The minimal toolset Is exhibited In Figure 5. The MAPSE depict- 
ed in the upper part of the figure consists of the services of a KAPSE 
(Kernel of the APSE) which interfaces to the other CAIS conforming MAPSE 
tools and the project data base. The six functional divisions of the 
MAPSE tools which sit on top of the KAPSE are: the compiler, the edi- 
tor, the linking leader, the command language, the debugging tools and 
configuration management tools. It is important to realize that the 
MAPSE toolset Is relatively independent of methodology and will provide 
the most help to the software engineer during the development phase cf 
the life cycle. 

Advanced technical tools are needed to reeforce specific method- 
ologies appropriate for earlier and later phases of the life cycle. Un- 
fortunately an Integrated set of tools appropriate for the life cycle of 
a project such as the space station does not exist at this time. Three 
steps are essential in order to meet the 1990 goals mentioned earlier. 

The Steps 

The first step is to establish a minimum baseline to begin the 
development of the system and software support environment. This would 
include a mature CAIS conforming MAPSE. The second step is to identify 
and select appropriate methodologies for all phases of the life cycle. 
This Includes methodologies for identifying and developing those reuse- 
able components which have a high potential for retum-on-investment. 
The third step is to build the reenforcing tools which facilitate and 
reward the appropriate use of the methodologies. This step is divided 
into two essential parts. The tools to reenforce the technical method- 
ologies must be reasonably well understood before the complementary ad- 
vanced tools to support management can be developed. 

Slice of I.lfe 

Figure 6 shows a slice of the life cycle [3]. Essentially this 
view is applicable to any phase of the life cycle or any step within any 
phase. It begins with a representation of the problem (or a proposed 
representation) which is now ready to be transformed according to some 
methodology into a new representation which is closer to the solution of 
the problem. The representation is fed into the project data base and 
appropriate verification and validation techniques are applied to insur^i 
that the new representation conforms to the letter and the spirit of the 
previous representation. The vsrification and validation results also 
enter the project data base and feedback is used to fine tune earlier 
steps. The boctora of Figure 6 a".;plles this view macroscopically to the 
phases. Initial concepts of the problem space are transformed into a 


■^ - '^-.■^•^'', 


requirements analysis document. The requirements analysis document 
consists of two major divisions: the "shalls" which must be demonstrat- 
ed at acceptance test time and the "shoulds" which are vital to the life 
cycle success of the project but which cannot be dichotomously demon- 
strated on a pass-fail basis at acceptance test time. 

The "shalls" of the requirements analysis document are trans- 
formed by an appropriate methodology into a design specification docu- 
ment. The design specification document specifies in annotated, com- 
pilable Ada package specification form: what processing is to be de- 
monstrated at acceptance test time, how well the processing is to be 
performed and under what circumstances the processing is to be perform- 
ed. The design team can then consider the various design alternatives 
which would meet the design specifications. The design alternatives can 
then be evaluated in the context of the "shoulds" of the requirements 
analysis document. A proft^ssional engineering judgement is used to 
select the best design alternative. 

Application of the Slice of Life 

The model of the slice of the life cycle can also be applied in 
a more microscopic sense to steps within the various phases. For ex- 
ample, the bottom half of Figure A shows an expansion of the require- 
ments analysis phase for large complex systems. Each of the four over- 
lapping activities can be subdivided into four parts: analysis, parti- 
tioning, allocation, and synthesis [2]. 

For example, the first activity can begin with an analysis of 
how many types of local area networks are appropriate to accomplish the 
goals of the space station program. This may result in a first pass at 
partitioning the computing requirements among local area networks to be 
located on the ground, on space station, on free flying platforms, and 
on other types of local area networks. As shown In the figure the 
principal requirements analysis team can continue their work while a 
selection of task forces can now begin the analysis of their particular 
LAN assignments. One team may analyze the LAN requirements for space 
station while other teams analyze the requirements of other entities. 
Each of these task forces will eventually reach the point of partition- 
ing their work Into various computing requirements. 

The task force focusing on the space station may determine that 
a fully mature space station '/ould have "N" cluster, of computing re- 
quirements (eg, a guidance, navigation and control cluster; a health 
maintenance facility cluster; and others). As a result of this parti- 
tioning, new task forces can be assigne 1 additional parallel activities 
to begin a more detailed analysis of the requirements for each of these 
computing clusters. At some point the further partitioning of the 
requirements of each of the clusters is analyzed in the context of the 
whole program so that common denominators can be factored into cluster 
components such as Network Communication Services, Network Information 
Services and Network Application Services. 

The second phase of the life cycle is macroscopically shown at 
the bottom of Figure 6 as the second transformation step which leads 




from a requirements analysis document to a design specification docu- 
ment. However this phase can be subdivided Into a more microscopic 
expansion of the design specification steps as shown in Figure 7 [1). 
The second step of this phase is particularly important to good systems 
and software engineering. This step produces a representation referred 
^o as the abstract functional specification. A characteristic of a good 
abstract functional spec is the separation of engineering concerns that 
allows the more difficult Issues to be analyzed with respect to risk and 
the probability of impact by future change. 

Those "dark" Issues identified co be "at risk" can be Identified 
in the project data base and can quickly be assigned to feasibility task 
forces. For each of these issues, the feasibility task force should 
begin by doing a saturation study. The purpose of the saturation study 
is to determine if a solution to the issue has been found by ethers and 
if the solution is documented and accessible to the tesm. If the team 
can convince their peers that they now know how to solve the problem, 
the issue may be removed from the "at risk" list in the project data 
base. The feedback from the project data base then allows this issue to 
progress to the next transformation. In many cases, however, the issue 
identified will not have been solved in the same context as the current 
application or there is Insufficient information available. 

The team may proceed to the consideration of detailed scenarios. 
It is possible that structrued-walk throughs of detailed scenarios may 
succeed in convincing the team and their peers that the issues are suf- 
ficently understood to be removed from the "at risk" list. Unfortun- 
ately a number of these walk throughs succeed in Improving the team's 
understanding of the problem but not in convincing them they have a sat- 
isfactory solution. 

^'^JL}- ^ Pr o typing 

It is then appropriate for the task force to build a rapid 
prototype. The rapid prototype is not a rapid prototype of the entire 
system. Instead it focuses on the separate concern which was considered 
to be at risk. If the understanding that emerges from work with the 
rapid prototype convinces the team and their peers that the issue can 
now be removed from the "at risk" list then the next transformation step 
can begin. Upon some occasions a rapid prototype may be a less appro- 
priate approach than a simulation. Even more frequently the rapid 
prototype may require interactions with simulations of the remainder of 
the environment. This is an area which is not well understood at the 
present time. 


To summarize the pracceding nectloas, a computer systems and 
software engineering support environment appropriate for 1990 Implies a 
mature, CAIS conforming MAPSE has been extended to provide the follow- 
ing: an Integrated and highly automated support environment consisting 
of a life cycle data base, a project library with a large collection of 
reusable components, good configuration change management, advanced 
technical tools, and advanced management tools for use with distributed 



•.:«M.i;':^^-*\ -.../■■ . :. ' «^ 


hosts, distributed targets, and an Integration, verification and valida- 
tion host with testbed. 


This work has been supported in part by NASA Contract Number 
9-17010. In addition to their colleagues at NASA Johnson Space Center 
and the University of Houston Clear Lakra, the authors gratefully ac- 
knowledge the support of the participating contractors and manufacturers 
which make the work of the Joint NASA/JSC UH CL APSE Beta Test Site pos- 
sible . 


The NASA/JSC UH-CL APSE BeU Test Site was established in Sep- 
tember 1983. The test site team has grown to 25 organizations supplying 
76 researchers. There are 64 tasks underway. The task order for the 
teams is titled: 

"Research in the Application of the Ada Programming Support 
Environment to the Life Cycle of Large, Complex, Distributed 
Computing Applications' 

The purpose is stated in the task order as: 

"The research will address the applicatioii of the Ada Programming 
Support Enviroument to large. Complex, distributed computing ap- 
plications (such as the Space Program) with a 

long schedule of modular growth 

(eg, 10.. 30 years) and a 

longer life cycle 

distributed hosts 

distributed targets, and an 

integration, verification and validation 

host and test bed. 
In particular the research will focus upon the issues which are 
not well understood (le, 'dark' Issues) In the computer systems 
and softwar'j engineering of such applications. The goals of the 
research are to 

identify and 

illuminate these 'dark' areas and to 

reduce the areas "at risk'." 


Terse definitions of space station compftlng terms tre listed 

Ada* Augusta Ada Bryon, Countess of Lovelace, daughter of Lord Byron 



.^^■:\%^^\ ^- ■ {«#■ 

waf) the assistant and patron of Charles Babbage and worked on 
his mechanical analytical engine. (The world's first program- 
mer). Also the high level language of the U. S. Department of 
Dfifense. *Ada Is a registered trademark of the U. S. Govern- 
ment Ada Joint Program Office 

APSE Ada Programming Support Environment 

CAIS Common APSE Interface Set 

GAN Global Area Network 

ISO/OSI Lnternatlonal Standards Organization Open System Interconnect 
(The seven level network communications model) 

JSC Johnson Space Center 


LAN Local Area Network 


NAS Network Application Services 

NCS Network Communication Services 

NIS Network Information Services 

RAN Remote Area Network 

UH CL University of Houston Clear Lake 


Henlngor, K. L. , Kallander, J. W., Shore, J. E., Pamas, D. L., 
Software R equirements for the A-7E Aircraft , NRL Memorandum 
Report 3876, 1978, 

Marlanl, M. P., Palmer, D. F. Tutorial; Dlatrlbuted Sys te m Design , 
IEEE Computer Society, 1979. 

McDermld, J., Ripken K. Life Cycle Support In the Ada Environment , 
Cambridge University Press, 1984. 

Ross, f). T., Goodenough, J. B., Irvine, C. A. "Software Engineering: 
Process, Principles, and Goals", Computer, May, 1975. 

Charles W. McKay 

Professor and Director of the High Technologies Laboratory at the Univ. 
of Houston Clear Lake. Dr. McKay is responsible for directing the 
research and development efforts of the Joint NASA/JSC UH CL Ada Beta 
Test Site. The test site team consists of approximately 25 companies 
and 75 principal Investigators. This team is researching the applica- 
tion of the Ada Programming Support Environment and Software Engln3erlng 
Principles to the NASA Space Station. 

Rodney L. Bown 

Associate Professor and the Technical Coordinator of the High Technolo- 
gies Laboratory at the UH CL. Dr. Bown is responsible for coordinating 
the technical activities of the Joint NASA/JSC UH CI- Ada Beta Test Site. 






WW( IIAIitai fi 

rRoroaro icrKKaiOfl ro« 

%1 •<. Cfcarlas «. Mcisf 



■ It. «..» 




— ft> 



Mil A*!^ 




•> •• 



Kl< ••»-• 





• It* 

tl/ac* Tr»H 


*/««(f l«ar 

rvML or cofWTCii mnns aw 


J IMIIVICI J ^.OtlllKt V,< 1) ' \fM\u)* 



•nuMtxiK . nojicT mi* utc 


riM«c 3 


(noroiio Ai A iTir iomuo 




BN«l*««flMfl IM •■(•lkt(S»U>«N 

• Md 

Mfl« •( ••«)«« 


1*« »tlMlvl«a, BCtlM** 



• !• 

■wpr«L' » 

t«**«M*l*« ■4tl«ll 111 

• €•#■•«* 


• t((l«l««t 

• W«»««f*(»*4«»l* 


1 JhfwhMi uif ui; u&JU •< ^>^* 



•IMW FMAJtt «r TNC WmOC Lift CYtUi 

!<«•« tk« ^ityMkl** •! 1 

»««tl*^*i Catfc«««ft 

• »«^li*M#l« *«*lial« Itt 

• »«alt*i •pMlll««tl«*« !•« 

• Mclan III 

• MvvlBfMAt lit 11% 1 

• Vflil(ic«ilMi •««« fall^atla* 

. Mmlwl* lc««l 111 

. I*tt«(*ll«« 1*1 

• H«Ul«H«M« %mt O^rstlMi 1 ••! 1 

•■ ""'^' — 

tN » 

. ^i- 

■AH • UU 

I *" 

- ► 

.... ^_ 


(>• rwoicTi 


(- i. ifty 

: »H ► . «i . • 1 


Ul AllacjilM 

s 1 «r«tk««it 












LllUltY /tu 



• tor n<A>f 

•HO»l.l • 



• tlCMHIC*!. 

• nuueinwr 



- DlimiiuTCO 


. iiirts"ATic 
Hotr t rttriiD 


'SLICC Of Ll«"(«<»i »fcBl««lB 4 lirafc. 19M) 

WAIIt 1(2 (lAWWU) 



Fcasibllty Taik Forcvt 

r«r ••MpJ*. l« ■<««« « MJ««f«M /art 
J ^liif *i** MJU«M«»ar* I 


-> HM.;«cT M!A MM 



S^>oop.Q"'-" N86-15178 


R. M. Mordlund, S. T. Vost. A. K. Woo 

Rockwell Int*m«tionaL/Rock«td]m* Diviaion 
Canoga Park, California 


The Rocketdyne Division of Rockwall Intamational haa anibarked 
on a productivity improveiMnt progr.ui in anginaering. Thia affort 
included participation in the White Collar Productivity laiprovenent 
(WCPI) project sponsored by the American Productivity Center. A number 
of things have been learned through this pkoject. It oaems that any 
productivity improvement project should be eaployea driven. The 
Rocketdyne project was essentially started as a result of a grass- 
roots effort to remove some particular hindrances, and employee en- 
thusiam was a prime factor in the continuing progress of the effort. 
A significant result was that awareness of problems at all 1 ^la in- 
creaaed. Many issues surfaced in the diagnostic phase, and were then 
noted and discussed. This process added legitimacy to issues that had 
previously been merely ur.spoken conc«*me. The initial foelings of 
many members of the pilot group was that aignificant chfingea would 
occur relatively quickly. It is now recognized that this will have to 
be an ongoing, long-term effort. A wa a Uh of i ns i ght is e o ntaii ied C ^ n 
t h e pil e fc gpou p t — end-iCBeping all those people involved ensures their 
rnntinuint r niil i llnirlTrrt t^tV inrrnadllhr— Thffli' twh rrndviBtivityi An added 
benefit to keeping a large number of people involved is that as 
changes are proposed an(? tried, there is increased accept.ince because 
those affected have contributed to the clianges. Thus, it is evident 
that this degree of employee involvement has made Rocketdyne* s program 


The Rocketdyne Division of Rockw xl lutemational is a conipany 
of approximately 7000 employees. The primary products of this division 
are liquid rocket engines. Rocketdyne designs, develops, and manufac- 
tures low volume, sophisticated rocket engines and related technology 
products that are often in the public eye. This type of company and 
this type of work commands an engineering staff comprised of dedicated 
and very capable individuals. 

American business and culture has entered into an ora when pro- 
ductivity in the work force is in the forefront of its attention. 


\ S^ 


Costs of doing business, and staylnft In business, a.:e high and will 
only increase. The primary survival technique Is to become as produc- 
tive as possible. This Is especially Important at Rocketdyne, where 
expensive and low volume prc'Mcts are sold almost exclusively to 
government agencies that recently have gained a renewed cost contain- 
ment focus and are willing to change contractors to spread out the 
available funds. 

Rocketdyne, along wfth other members of the aerospace Industry, 
was hit extremely hard in the early 19708. Companies were forced to 
drastically cut back by reducing the size of the facilities and thd 
number of employees. A few years passed while regrouping and working 
to develop new products. Due to key contracts such as the Space 
Shuttle Main Engine, Rocketdyne began to grow and expand. As a result 
of this growth and expansion, a large fraction of the present engi- 
neering ptaff is young. Th. ^ has created an interesting bimodal work 
force of older employees vfio were retained throughout the earlier lay- 
offs and the newer, younger employees. 

The productivity influence in this environment manifests itself 
differently on individual employees. Changes to increase productivity 
are generally changes to the traditional or comfortable way of opera- 
tion. Resistance to change varies between people, but typically in- 
creases with age and/or experience. Thus, the bimodal employee age 
profile stratifies employees in response to change. 

In addition to this employee stratification, there are signifi- 
cant differences in personal insights, values, cultural background, 
job history, and work ethics. The combination of stratlficatior. along 
different lines and simpl*' personal differences has created a nonhomo- 
geneous engineering staff that is not bound to traditional lines of 
thought. The lac of continuity throughout the engineering rtaff is 
reflected by a lacK of subsci'lption to the company line. In light of 
the industry requirement for a more productive work foc>:e, this cross 
section of employees has been primed for response. 

The roost straightforward means toward increased productivity in 
the work force is in the manufacturing or production area. Because of 
this, eiuployee Action Circles (Quality Circles) were fostered and de- 
veloped among the manufacturing personnel . These groups were directed 
toward problem solving and streamlining the m;:.nufacturing process. 
Realizing the relationship betw«>,en good engineering and a smooth and 
efficient manufacturing process, the idea oT similar problem solving 
groups within the engineering department war. suggested. 

in May of 1983, a first line manager in the Engineering Design 
Technology Department was asked to coordinata such a group with the 
help of a Quality Circle facilitator. Because Employee Action Circles 
were viewed as a blue collar activity, this white collar group was 
called a Nominal Group Technique (NGT) group, after the methodology to 
be used in the problem ioent; firacion and solving process. Involve- 
ment with the group was voluntary, and invitations were extended to 
the members of the manager's unit. The purpose of the group was to 
look at ways to increase the employees' productivity. 

News of an employee group formed to legitimately raise and 
address comnon productivity concerns spread throughout the engineer- 
ing department. There were numerous productivity sore points among 
the engineers, and an MGT group ma viewed as an opportunity to have 
a say in making th^ organization function more effectively. The 
timing was fortunate, for there were lingering doubts about manage- 
ment initiated productivity programs that had recently been at- 
tempted. Earlier activities, gueh as an intimidating personal time 
usage study summarized in pie charts and productivity increase news 
bulletins the engineers were instructed to write on a quota basis, 
had left a bad taste among the engineering staff. An NGT group ap- 
peared to be a way engineers could creatively address and identify 
productivity obstacles, and then tfurk toward thoir elimination. 

More than a way to help the company become more competitive, 
profitable, or more secure in its business sector, interest in in- 
creased productivity was seen by the engineering staff as a way to 
eliminate or reduce frustration and obstacles standing in the way of 
doing the challenging and interesting work they were hired to do. 
Increased productivity would mean the ability to be free to perform 
at a high level, and not be distracted or hindered in that pursuit. 
Ttiere were personal goals to attain, so interest was high. 

The first NGT group completed its activity with a summary 
report in November of 1983. That report addressed such issues as 
unnecessary analyses, organization and retrieval of records, noisy 
working areas, and the avenues pursued to initiate a favorable 
change for each iseue. During that time period, two other groups 
were allowed to form from an individual unit with the direction and 
help of the first-line manager and a facilitator. Those groups 
worked on various problems through approximately June of 1984. 

The NGT groups were focused at short range, addressing pr<~.- 
lems affecting efficiency and morale with much energy and emotion. 
Many of the issues were met with responses from management that 
there was no budget to Inclement the suggestion, or that it was con- 
trary to the present coiipany operating policies. To a degree, these 
responses confirmed soroci of whe employee frustrations with the sys- 
tem, and they deepened some resentment toward management's attitude. 
NGT groups were authorized to meet and submit recommended changes, 
but there was no manag<:sroent commitment to respond. There was no 
overall plan or direction to the NGT program, resulting in very 
limited goals being achieved. 

In May of 1984, the American Productivity Center (APC) con- 
tacted the Rockwell corporate Productivity and Planning Department 
regarding the White Collar Productivity Improvement (WCPI) Program. 
The WCPI, a research project conducted by the APC, was one of the 
programs selected by Rockwell to attack the productivity issuea 
among its white collar employees. The Rocketdyne Division was chosen 
to be the site of one of the WCPI pilot groups. It was then decided 
that tho pilot group at Rocketdyne would be off to a head start if 
it utilized an employee base that had already demonstrated an 


•- - ''3*V*"^. 

interest in increased productivity, so the pilot group was centered 
around the groups with HGT experience. 

The director of the Design Technoiogy Department at the 
Rocketdyne Division endorsed the project and introduced it to the 
pilot group. He was then appointed to another position within the 
division in the beginning of the project, and the departnent was 
left under the direction of an acting director. In retrospect, the 
project was somewhat hankered by the absence of a full-time depart- 
ment director. Some lack of authority from the top was felt When 
decisions trare required to expedite changes. The pilot group at 
Rocketdyne, comprised of about eighty people divided into six units, 
performs analyses that support all of the programs in the division. 
This pilot group is one of the biggest in the APC project. 

The UCPI methodology consists of six phases: Diagnosis, Ob- 
jectives, Measurement, Service Redesign, Team Development, and 
Technology Parameters. Ttiis approach is different compared to the 
^.raditional productivity improvement programs that concentrate on 
increasing output, which are more appropriately used in a blue col- 
lar, production-iype working environment. The trtiite collar working 
environment involves a unique interaction between quality, effi- 
ciency, and effectiveness. The WC?1 methodology provides a logical 
framework for an organization to review and enhance its operations. 
An organizatiot. looks at its functions, determines the reasons for 
performing its services, 5ets feedbacks on the performance, finds 
ways to improve its services, and assembles the people and equipment 
necessary to accomplish the tasks. 

The human dynamics aud operational practices of the pilot 
group were fjrveyed in the diagnosis phase. This provided a baseline 
for the project. Everyon*. in the pilot group filled out a survey 
questionnaire developed by the APC. One-to-one interviews of a cross 
section of the pilot group members and their users were conducted by 
a neutral party. This survey showed that the pilot group provides 
state-of-the-art technical analyses, is highly motivated, and enjoys 
the technical challenges. Three sensitive areas that the pilot group 
as a whole felt could be improved were the working environment, 
lateral and vertical conmunications, and efficient u«e of computing 
equipment and peripherals. This survey did not uncover any issue 
that was unknown to a majority of the pilot group, but it documented 
the concerns that were mere vevbal complaints before. Although the 
project was viewed with skepticism, a volunteer steering cnnssitlee 
for the project consisting of botn first-iine mansge^rs and technical 
staff was formed. Involvement- teams were also formed to attack the 
three previously mentlc.ied productivity hindraiice areas. 

Ttie diagnosis phase set the stage for this project. The com- 
ruittee and the involvement teams got people involved in the begin- 
ning of the project, and they were free to explore solutions. The 
involvement teams proposed and effected s^.,me changes within the 
pilot group. The resultant changes were accepted readily because 
people affected were directly involved in formulating the changes. 

:^ The six units reviewed their functions during the objectives 

t phase, and then determined the purposes for performing them. This 

■' effort resulted in the purpose and mission statement of the entire 

department. Every unit's manager determined the main services per- 
formed by the unit, «nd the staff in the unit brainstormed the ob- 
~ jectives of the services. An N6T method was used to rank the objec- 

J tives of each service. The service objectives were then revletred by 

I the users. 

Going through this phase gave everyone in the unit a better 
; sense of direction and perspective. This exercise rhowed people how 

they fit in the whole organization, and how their contributions 
^; affect the progress of the tasks. This phase took more time than 

■■';' anticipated to complete. Much time was spent in discussing why cer- 

tain services were performed. The process could have gone quicker 
if the manager had set the objectives also, and then had them re- 
^ viewed by the unit and the users. After all, the manager should set 

^ the goals for the unit. The objectives are a good tool for managing 

.1 and prioritizing tasks performed in the unit, and they are espe- 

,% cially useful for indoctrinating new staff. The review by the users 

.« <*'as essential because it f.ave the unit a sense of the expectations 

^ placed on them. The process promotes communication between the users 

t' and members of the unit. 



- Measurements were formulated to measure the service objec- 

z'. tives. It was drsired to develop measurements that are simple to 

^- make, not tiae consuming, and not traceable to any individual. The 

' x, steering committee felt the me&surements should consider efficiency, 

-^ effectiveness, quality, timeliness, employee attitude, motivation, 

;' resource usage, and user satisfaction. The objectives from the six 

^ units in the pilot group were combined to form a list of common 

group objectives. Heasurements were formed to measure these common 
pilot group objectives, so less time would be required by each unit 
to develop and make the measurements. The project stagnated somewhat 
at this point because the steering committee was puzzled over the 
appropriate measurements to be made. A lack of expertise to formu- 
late the appropriate measurements was felt. A separate productivity uvement project within Rocketdyne had an expert in measurement 
on staff, and it had taken them one year to deve " and validate a 
^, survey of eight questions. With the completion of this phase 

approaching, the steering committee decided to ^i "> ahead. They 
took the information at hand and derived a user survey, an analyst 
survey, and a list of measurements for the individual units. The 
measurements were reviewed by the units, so each unit could tailor 
the measurements. 

Measurements are usually the stumbling block of most white 
collar productivity projects. Individuals are generally wary an'i 
reluctant to submit to measures of their productivity. The steering 
committee was concerned about how inanagement would use the measure - 
ments obtained. Recognizing that there is uO perfect list of mea- 
'' surements, the steering committee was deemed to be the most quali- 

\! fiefl to formulate them. Familiarity with the operations of the pilot 

group helped formulate the measures. The questions in the surveys 



■'Si ^ 

were similar to the questions in the other productivity improvement 
program survey. This vas a otartins point from Which the neasuLes 
can be refined at a later date. Effort is in prosress to obtain the 
formulated measurements. 

Bach unit constructed flotrr.harts depicting the my its ser- 
vices w»re perfomed in the service redesign phase. The hindrances 
became visible, so they presented opportunities for iaprovaaent. 
The hindrances from the units were combined to form a list of conmon 
group bottlenecks that could be attacked collectively. Some service 
changes to f' reamline operations were proposed to nanagenent. Ser- 
vice redesign is an ongoing process, since there are always new sit- 
uations that present new ways to operate more efficiently. 

The Rocketdyne White Collar Productivity laprovenent project 
is ongoing. The above sumnary describes the first pass through the 
APC methodology. As a result of the enthusiasm generated by the 
Design Technology pilot group, three more pilot groups within 
Rocketdyne were formed. Other units within the Division have shown 
interest as well. Rocketdyne 's desire is ,o make the WCPI methodol- 
ogy a continual process. 

There have been a number of changes that have occurred as a 
result of this productivity effort. One of the most significant 
changes at Rocketdyne has been an increased awareness on the part of 
employees and all levels of management. In many cases, problems 
that were uncovered had been in existence for a long time. The pro- 
cess that tras use! during this program legitimized these issues, and 
created a framework for discussion and resolution. Although this 
process can occur at any time in an organization, a directed effort 
such as this encourages the kind of thoughts and exchanges bettreen 
members of the staff that often leads to resolution of the 
problems. Awareness seems to be a real key in creating the kind of 
climate in which productivity improvements can be considered. 

Another key to the success of any white collar productivity 
improvement project is that the project roust meet the ne&t^s of the 
employees and be desired by them. Concurrent wit:h this, however, 
there must be full management support. This is not to say that man- 
agement must be heavily involved in the process, or directing the 
effort. Any such project that is so overshadowed by management that 
the employees do not feel a part of it will be only marginally suc- 
cessful at best. An effort driven and sustained by the employees, 
on the other hand, will by definition have the energy and interest 
necessary to get the best return for the time spent. Since each 
individual manager perceives the increased productivity benefit to 
his area differently, support will never be uniformly avt^ilable. 
Some managers will even be openly hostile. But as an effort pro- 
gresses, opportunities will become available for these people to be 
convinced of the value of the project. An effort should be made to 
communicate all of thf^ successes and improvem.ents that occur along 
the way, >)ecause success breeds interest. The continuing success of 
the projt will be dependent on getting as much management support 
as possible. 




or:c;':a:. f.\se b 
of poor quality 

A fairly coimon fallacy In many organitations Is that pro- 
ductivity iapfovaatent is a discrate type of effort. Protrams with 
fixed tinM periods ara often established in the hopes that perman- 
ent, beneficial changes trill occur. It seeas that a mora realistic 
view is that productivity iaprovaaent needs to take place in an en- 
vJLronnant of onsolns effort. The very nature of the process, includ- 
ing the iaportant role of heightened awareness, is one of long-term 
coonitaant. In this kind of environnant, problems and iapediments 
to efficiency are constantly being ferreted out. Solutions to these 
probleaks are constantly in work. In the ideal case, the search for 
better ways to do things becoaes part of the organixat ion's standard 
operating procedure. At Socketdyne, the idea that productivity iin- 
proveoent is an ongoing effort has given individuals the impetus to 
continue working on probleas for which there are no short-term solu- 
tions. It also erased soae of the l>ad feelings created by past 
efforts that strove to obtain significant productivity iroprovements 
without a sustained el fort. 

In any organisation, the most important assets are the 
people. The knowledge of how things work, and how they could work 
better, resides in these people. They are the collective wisdom of 
the company. It is iaportant to let these people contribute to the 
solutions of theit own problems. In many organizations such as 
Rocketdyne, people are highly motivated and desire to do their Jobs 
.^s c/topletely and efficiently as possible. They can be expected to 
approach the opportunity to improve things relating to their jobs 
with enthusiasm. The wealth of insights into the workings oJ the 
organization is a resource that should not be wasted. In addition, 
keeping these people involved helps to ensure that they will accept 
the changes that result from their effort. 


Rocketdyne applied the American Productivity Center methodol- 
ogy to a pilot group in the Engineering Department. The methodology 
provided a framework for the establishment of an ongoing effort to 
improve white collar productivity. In addition, the effort has pco- 
vided an outlet for employees with a legitimate desire to improve 
procedures affecting their jobs. Another benefit in having an on- 
going productivity project was that many lingeriiig but un.-jpoken con- 
cerns became legitimate issues. Extensive inputs from users were 
obtained. Communication at all levels was enhanced. Work on improv- 
ing productivity will continue, but there have been accomplishments 
to date and a system has been put in place to facilitate future 



. # r~-_ - 




R. M. NOROLUIID - vas scA<luated from the University ot. Washington 
located In Seattle, Washlnston, in June 1.980 with a B..'>. in Hechan- 
Ical EnglneerlnS' Since that tine he has worked in the Aerotharmo- 
dynanlcs Department at Rocketdyne, and is now a Lead Bngineer tfork- 
ing on Space Shuttle Main issues. 

S. T. VOGT - received both B.S. and M.S. degrees in Mechanical Sngi- 
neerlng from Purdue University. He has spent the last three years 
at Rocketdyne in the areas of analytical heat transfer and experi- 
mental aerodynamics. Prior to that, he spent three years at Hughes 
Aircraft Company in the Missile Systems Group. 

A. K. WOO - received his B.S. degree in Aeronautical Engineering 
from California Polytechnic State University, San Luis Obispo. Hie 
position at Rocketdyne is in the area of rocket engine perfomance 
analysis. Prior to joining Rocketdyne, he was a student trainee in 
Flight Testing En'^ineering at Edwards Air Force Base. 








America's quality of life Is attributed by many to the technological 
advances made possible by our scientists and engineers. In the ! 980's 
factors are occurring to force U.S. Industries to recognize that: I ) our 
engineers and scientists are a critical resource and 2) this resource Is not 
being used to Its full potential. America's Industrial giants are experimenting 
with many approaches to Improving productivity in manufacturing but are 
still mainly "wringing their hands' regarding engineering organizations. 

As an internal consultant at Motorola, Inc. this author was selected to 
investigate methodologies to Improve engineering productivity. This paper 
will review the rocky road to Improving engineering effectiveness utilizing a 
specific semiconductor engineering organization as a case study. The 
organization had a performance problem regarding new product introductions. 
With the help of this consultant as a "change agent" the engineering t^am used 
a systems approach to through variables that wer^ effecting their out put 
significantly. This paper will discuss the critical factors for Improving this 
cnnineering organ I zat left's effectiveness and the roles/responsibilities of 
manageinent, the Individual engineers and the internal consultant 


Motorola, Inc., an International leader in commercial and industrial 
electronics, generated $5.5 billion In 1984 sales. The company, headquartered 
in Schaumburg, Illinois, manufactures a wide range of eicctronic equipment. 
Products Include systems and components ranging from cellular radio 
telephones and two-way FM radios to data communication products and 



-- r .^•^rr-' 



Approximately 100,000 employees work for five distinct operating 
segments In 1 10 countries. Each segment, known as a sector. Is divided Into 
groups or divisions depending on the size of the organization. 

The Japanese have captured an increasing segment of many Industrial 
markets In the last 20 years. These Include steel, consumer electronics, 
automobiles and hardware. The Japanese have targeted three specific 
markets which Motorola sen/es directly or Indirectly: semiconductors, 
communications and computers. 

In 1983 the Semiconductor Products Sector of Motorola, Inc. identified 
engineering productivity as a critical strategic issue facing the organization. 
An Executive Vice-President assumed the role of champion. A sector-wide 
committee was formed to gather data on productivity issues and make 
recommendations. After the problems were identified and prioritized the 
committee recommended that a manager be assigned full-time to address the 
issues. In 1 984 Janet Flero was selecte<i as Manager of Engineering 
Productivity. This decision was based on her success at developing the 
corporate Total Quality Improvement training strategy. 

Approximately 60 engineers and engineering managers were 
Interviewed resulting in several projects using different approaches. The 
project Involved with improving a division's new product Introduction record 
was selected as the basis for this paper. This case study best depicts the 
potential success of such an improvement effort. 


Recognition of the Problem 

The vice-president of a particular semiconductor division began having 
a series of meetings to improve the division's new product Introduction 
record. Three Product Managers, a Marketing Manager, a Planning Manager and 
a Manufacturing Manager reported to this V.P. Design Engineering reported 
into the Product Managers. Through a long and Introspective process the 
management team explored why the current level of performance relative to 
new products was not meeting their expectations. The initial tendency was 
to spend their meeting time trying to find individuals responsible for the 


• . *•■-.%. 


slippage In schedules. The management team resisted this tendency to 
"search for the quilty" and looked for the factors that had changed relative to 
their previous success record of new products. 


—Circuits had gotten larger, exponentially larger, and resources had 

Increased linearly. 
—Check and balance system had slipped. Resources had left and had not 
been replaced 
•, —Criteria for successful prototype had become much tighter. 

—Allocation of resources had shifted from new products to existing 

products (I.e. from long term to short term). 
—Demand for new markets runs counter to economic times. In good 
times like 1984 the demand for new products was less. Internal 
,. J priorities focused on "milking" existing products rather than 

^ : developing new ones. 

^1 —Rapid growth of the division had resulted In people who were 

J. inexperienced at their current level. Design Managers In particular 

;I had moved rapidly from "hands-on" work to the management of the 

design process. These people needed effective, job-related training. 
■ . —The design organization had been decentralized and flexibility to 

•:, allocate resources for special projects like CAD applications was 

—New CAD and project management tools were being developed by 
central research but existing tools were not being maintained or 
—The number of Interfaces for the design engineer had Increased. 

The matrix structure of the organization was more complex. 
—The organization had become more internationally oriented. 

Analysis of the Problem 

Introduction of the concepts of the Performance System helped the 
vice-president and his direct staff to step back, gain perspective on their 
problem, and reach some appropriate actions. This Performance System was 
developed by Dr. Geary Rummler of Summit, New Jersey. The system is based 
on distinguishing the difference between the performer and the perform awc^. 
The diagram on the following page Illustrates the principles of the 
4 Per Tormance System. 

•i' 241 







•"—7 P»r former 



Figure 1 The Performonce System 

It was imperative for the top management team to recognize that poor 
performance of the Design Engineering Department regarding new product 
introductions might have resulted from a breakdown In any of the five 
components of the performance system. In one meeting the V.P. and his direct 
staff reviewed the following questions relative to the 5 components; 

\. The Output 

—Is there adequate and appropriate criteria (standards) with 

which to judge the design team's performance? 
—Do the Design teams l<now what is expected of them? 
2. The Input 

—Is there a clear cr sufficiently recognizable indication of the 

need to perform? 
—What interference is there from incompatible or extraneous 

—Are the necessary resources (budget, personnel, equipment, etc.) 

available to perform? 
—Are the products adequately specified and defined for the design 


J 42 


3. The Consequences 

—Are there sufficient positive consequences (Incentives) to 

—Can we eliminate the negative consequences (disincentives) to 

4 The Feedback 

—What kind of feedback exists ss to how well (or how poorly) the job 

Is being performed? 
—Is the feedback timely and appropriate? 
5. The Performers 

-Do they have the necessary knowledge and skills to 

perform the job? 
—Do they have the capacity to perfom-i both physically and 

—Do they have the willingness to perform (given the incentives 


From this analysis the division management team decided the output 
most critical to their ongoing success was to design products with much 
fewer reworks or "passes." The team recognized they had been focusing the 
organizations priorities on today's profits rather than tomorrow's products. 
This appeared to make the design group's performance less important. They 
admitted that inadvertently they might be rewarding rework by giving 
priority and attention to those who were most delinquent. There was little 
tracking or feedback about the status of various products and the number of 
passes that had occurred. With the realization that the performance system 
was breaking down In a number of places the management team "rallied 
around the flagpole" making First Pass Success the overriding goal for 1985. 

Earlv Actions by Management 

1. The Output— lOOX First Pass Success goal on new products was made 
public to all members of the organization. The 1985 goals reflect'»1 goals 
that supported th*f jperordinate goal Some delinquent products were 

2. The Input— Two areas which received Immediate attention were 
marketing plans and CAD software packages. Both of these areas 
required the clout of the V.P. to Improve the quality of the service. 



- '|P r*^- 


3. The Consequences— Incentives were Instituted for designs that functioned 

ors the first pacs. Attempts were made to remove priority from delinquent 

product designs. 
4 The Feedback— Tracking systems were implemented. Design reviews held 

t'uring the development cycle were revised. A hjg»ier level of management 

began attending design reviews. 
5. The Performer— New expertise In CAD applications was added. Existing 

personnel were trained ;n Project Management Skills. Design l^-^r^agement 

'yeqan working with the Internal consultant to analyze the Performance 

System within the design englneerlrg group. 

Ur ly Ac tions t?y Design Managgrs 

"^Ive Design Managers :7iet to redesign the method? and procedures In 
which they created their outputs. The methodologies used do design 
yesterdays products were not working well ^oday. The des;:>'i managers 
moved through a process that allowed them the Insight to revise their 
procedures The events occured as follows: 

• Defined the output (First Pass Success) In terms of functionality, D.C. 
& A.C., electrical, manufacturablllty, arj schedule. 

• Identified Interim outputs and corresponding standards. 

• Identified their "customers* (I.e.the departments that received their 

t Negotiated with the "customer" to oetermine If they agreed with the 
standards of the output which would become their Input. For most t^l? 
was a first! AH department began to trust that if they received 
quality inputs they could meet their schedule committments. 

• Listed all the tasks necessary to produce the design outputs. 
Redesigned the sequence of the tasks for maximum effectiveness. 

• Redesigned the meetings that monitored new products. A process for 
technical de-sign and business trade-off decisions was Implemented. 
Cross-functional champions of new products were ld*^^tified a; id 
brought into the review process earlier to be part of the trade-off 

• Identified inputs to the design process, specified standards and 
communicated these to the appropriate departments. 

• Initiated and monitored checklists soecifying required check points, 
f Wrote individual Performance Appi .isals to include First Pass 

Success standards. 



• Requested that departments effecting new product Introductions 
' move through the same process as the design group. This process 

would define: what service or product they produce; what Inputs they 
needed; and how they woulo know If their "customer" was satisfied. 

Organizational Mapping 

The process used by the Design Managers was, an application of the 
; concepts developeJ py Geary Rummler entitled "Organizational Mapping 

Process". This process is based on the premise that organizations operate as 
systems, consisting of subsystems (functions/departments) requiring 
specific Inputs and outputs in order to meet the organization's (system's) 
objective. /\n organization's effectiveness Is usually a function of how weii 
% J these organizational subsystems are llnKCd or connected. System 

^ "disconnects" are frequently at the root of organizational performance 

"I issues. 

* Org,iilzation Mapping is an effective and efficient process with a 

visual format which represents an organization as a system and guides a 
management team through a focused analysis to effective action plans. A 
flow chart is the visual that d^^plcts h^w the inputs and outputs of each 
function Interface to produce the final product. 


-; The Oi u^nization Mapping Process has been used with teams ranging 

[ from a pres*'*^nt and vice presidt.^ts lo peer work groups addressing such 

cr I M c«< I bus 1 ness i ssues as: 

9 the Implementation of a new division strategy; 

• the Implementation of a division reorganization; 

• the Improvement of the product design process, 

"" f the Improvement of manufacturing yield, quality, cycle time, 

invAotory ,cost and delivery; 

• the merger of functions; 

• the design and start-up of new fun':itons. 

The Organization Mapping Process results in 
I • a cross-functional, comprehensive plan for addressing the 

I Critical Business issue. 

t the potent id I for a stronger management "team." 
■"■; f ? common process for addressing cross-functional Issues in 

'-; the future. 




.Jl.^^ /.\-^^ 


Project Hindsight 

Progress was made tn the organization described in this case study. In 
1984 only 12.5% of the new products were First Pass Successes. In 1985, If 
the remaining months hold true to the new trend, over 50% of the products 
designed will meet First Pass Success standards. That's the good news. The 
bad news is that the extra effort required to set up the systems and 
procedures to ensure First Pass Success has caused a 25% reduction in 1985 
new product Introductions. This case study Is a clear example of an American 
management team making a long-term Investment. The team almost buckled 
under due to the painful short-term consequences. Now they realize that the 
long-term gains will come. 

Hlnc^-ight shows the project could have been done more efficiently and 
effectlvelv. Looking backwards and learning from mistakes is not considered 
a worthwhile activity In most companies. For significant improvements to be 
mad*; we must accept and even reward critiquing. Rewarding the individual or 
group who uncovers a problem is much more productive than searching for the 
guilty and shooting them. 

The design engineering group did an excellent job of listing all the 
tasks that they "should" do In order to ensure First Pass Success. It would 
nave been helpful to make a list of all the other tasks they do that detract 
from the list of good Intentions. McDonnell Douglas Electronics Company 
launched an improvement project similar to this one. Two lists were created: 
value-added tasks and non-value-added tasks. A goal was made to decrease 
non-v?lue-added tasksde.waste). This effort, coupled with other 
Improv/ement activities, allowed engineering resources to decrease through 
attrition during a growth year. 

Th( • "Sign group at Motorola recommended to their interfacing 
departments that they map their organizations specifying the standards fo~ 
inputs and outputs. This action was followed through by some but not all 
departments. Thus the innovative improvement activities of the design group 
were not being reinforced by the larger system. The lack of division-wide 
buy- in to the new methodology left the design group feeling totally 
responsible for the performance though many tasks were outside tneir 





^!^: .: ;:.„, " ^ 


Many other factors affecting new product performance were !{jt*ntlf)ed 
and not fully resolved. These Included Improvements In: the pi oduct planning 
process, the goal setting process; engineering computing tools and support; 
and training needs assessment. 

The role of Interna! consultant In this project was awkward and 
ambiguous. A contract Detween the consultant and vice-president would have 
clarified the roles and expedited the project. To be effective an Internal 
consultant must nenr Jate with the client the following: 

• an understan Jing of the problem 

• the*scope of the project 

• the client's expectations 

• the consultant's expectations 

• agreement on deliverables and tlmefrarr ? 


External forces facing American Industry today demand internal 
changes. Change Is necessary but very difficult. Each of us has noted at one 
time or another " The cniy constant thing around here is change." Change is 
often a reaction to external factors. Change is not often enough due to a 
proactive choice. This case study shows an organization undergoing change. 
This change was painful. The Individuals atjvocating change were resisted, 
but progress was made. We all have filled our notebooks at conferences with 
innovative ideas that back in the "real" world fail in the implementation. The 
complexity of the change process must be appreciated before you venture back 
home to implement the pearls of wisdom received today. 

Suppose research validated a theory that people at an optimum weight 
are more productive then people who are overweight. You, as a conscienscious 
manager, request that the personnel organization select the best weight 
reduction program with all necessary training components. You agree to send 
necessary folks t'^' the events. The first problem may be to get the personnel 
organization to u,jrs.^e to help you. The second may be getting anyone to attend 
the "events', vqu can always nake them go, but, will the objective of weight 


loss be achieved? Most of us recognize we have areas that need 
Improvement-losing a few pounds, stopping smoking, or exercising more. Few 
of us do anything serious about It until a significant event, such as a heart 
attack, strikes us down. 

Must we wait for a corporate "heart attack" in our companies before we 
choose to implement change? Is reaction our only choice? Managing a crisis 
is easy. Managif>9 improvement is difficult. One Is reactive. One Is proactive. 
When managing a crisis the problem is well defined and the resources to be 
allocated are obvious. When managing improvement the problem and 
opportunities are numerous and ambiguous, plus resources scarce due to 
previous crisis allocations. The successful managers of tomorrow will be 
those who clean up Ineffective systems and create new systems that improve 
the output. CONTINUOUS IMPROVEMENT must become a way of life for 
companies to move from surviving to thriving. Improvement requires change. 
Change will never be popular. As you listen to these ideas and go home to 
advocate change recognize the task at hahd is not easy. If you feel resistance 
you are probably effecting change. Good luck! 




Janet D. Fiero Is a consultant to industry on how to manage for quality and 
productivity improvement. She started her business, Improving Engineering 
Effectiveness, Inc, In 1985. Her areas of expertise include Statistical 
Engineering Applications, R&D Organizational Aj,alysls, and Engineering 
Training Development. She has 5 years of experience *n training and 
performance Improvement manag.»ment at Motorola Inc. and 12 years 
engineering and management experience at Motorola Inc. and Falrchild 
Semiconductor. Janet graduated \\\ 1968 from Pennsylvania State University 
with a BS degree in Biochem<otr/ and is currently pursuing an MBA at Arizona 
State University. She is ? member of ASEE, IEEE and the Technical Education 




'«V 'V V, .-'X. V - VwM*/ 


Software Producii vUy Improvement Tnrojyn Software 
Lnyineeriny Technology' 

Frank E. McGarry 


It has been estimated tnat NASA expends anywhere from 6 to 10 percent of 
its annual budget on the acquisition, implementation and maintenance of 
computer software. Although researchers have produced numerous software 
engineering approaches over the past 5-10 years; each claiming to be more 
effective than the other, there is very limited i|uantitative information 
verifying the measurable impact that any of these technologies may have 
in a production environment. At NAiiA/GSFC, an extended research effort 
aimed at identifying and measuring software techniques that favorably 
impact productivity of software development, has been active over the 
past 8 years. Specific, measurable, software development technologies 
have Deen applied and measured in a production environment. 

Resulting software development approaches have been shown to be effective 
in both improving quality as well as productivity in this one 
envi ronment . 



f^^tfh ' 

\\' - ' ^ 


The Software Enyineering Laboratory (SEL) was established in 1977 by 
Goddard Space Flight Center (GSFC) to investig-.te the effectiveness of 
software engineeriny techniques as applied to the development of ground 
support flight dynamics systems. The goals of the investigation are (1) 
to understand the software development process in a particular 
environment, (2) to measure the effects of various development 
techniques, models, and tools en this development process, and (3) to 
identify and apply improved methodologies In the GSFC environment. ScL 
research should provide the knowledge to enable GSFC to produce better 
quality, less costly software. 

To accomplish these goals, the SEL studies software for satellite mission 
support during its development life cycle. This software is developed by 
the Systems Development Branch at Na\SA/GSFC, which is responsible for 
generating flight dynamics support software for GSFC-supported missions. 
The software includes attitude determination, attitude control, maneuver 
planning, orbit adjustment, and general mission analysis support systems. 

The SEL continually monitors and studies all Systems Development Branch 
software, which includes software developed both by GSFC employees and by 
contractor personnel. This data covers software development projects 
that started as early as 1978 and as late as 1985; and the SEL 
anticipates mat data will continue to be collected and studied in the 
future. Approximately 50 projects, which range in size from 200i; lines 
of source code to over IbU.OOO lines, nave been involved to date. 

While Investigating projects totaling more than 2.5 million lines of 
code, SEL personnel gained insight into the software development process 
and began to discern trends in the relative effects of various techniques 
applied CO the software projects. This report: 

Describes the -notivatlon and background of the SEL. 

Relates the concepts and activities of the SEL. 

Summarizes the results of SEL research. 

Reports the status, conclusions, and recommendations of tne SEL. 




Extensive efforts nave been made duriny recent years to devise iinproveo 
software development techniques. This work yenerated numerous tools 
(e.y., precompilers and proyramner workbenches), cost and reliability 
models, and methodoloyies (e.y., structured proyrammiriy and top-down 
desiyn); all were supposed to improve the development process. Early 
evaluations of the effectiveness of these techniques were incomplete 
and/or inconclusive. This may have been due, in part, to an unrealistic 
assumption that the software development process could be isolated from 
the environment in wtiich it occurs. However, no element of the 
development process can be understood outside the context of related 

For example, productivity in some environments may be constrained by 
staffing patterns. T!tus, the possible beneficial effect of a 
productivity-enhar,ciny methodoloyy may remain unrealized and unrecoynized 
because of an in?.ppropriate allocation of manpower. 

The :)EL approac*! to software enyineeriny research is nolistic. Its four 
components are a problem statement, an environment, a process or 
activity, and a product (software). The' development process is 
subdivided into seven sequential phases of activity. A yoal of SEL, 
then, is to refine ^ .e definitions of the model elements and to define 
their relationships. 

Tne first step toward this yoci is to und erstand the software 
development process currently in operation and its environmant. 
Important attributes of the software proDif:Ti and prod'jcts must also be 
irvestiyated. Such an uodersLandiny provides a baseline from which the 
effects of attempted improvements can be measured and allows the 
identification of streny!:ns and weaknesses so that efforts can De focused 
on the areas of yreatest need. 

Beyond understanding the current process and environment, the SEL is 
interested in improviny that process and environment. The SEL recognizes 
i four-step procedure leading to more effective software development. 
The steps are to: 

Become aware of the development techniques available 

Evaluate tne available techniques to determine those most 
ef fecti ve 

Engineer (customize) those "Dest" i.ecnniques. 

Tnjs procedure can beconie the basis of a mijular system of self- 
evaluation and improveinent , whereby ds new techniques become (aVoilable, 
they are tested and incorporated in the softwar., ievelopment process. 


■v^-v V'V' - ■ {:i} 


The overall objective of the SEL is to understand the software 
development process and the ways in which it can be altered to improve 
the quality and to reduce the cost of the product. However, the SEL has 
defined some intermediate objectives within the previously defined areas 
of concern that will help meet that general yoal . These objectives fall 
into two classes: experimentation and communication. 

Experimentation involves evaluating existing software development 
technolo'jies and developing new technologies. Specific ob.j&ctives of the 
SEL are to: 

Conduct controlled experiments 

Evaluate software development methoaologies 

Evaluate software development tools 

Analyze cost estimation models 

Analyze software reliability models 

Develop a set of software quality metrics 

The results of experimentation must oe Incorporated in the software 
developiiient process to improve it. This process requires communicstion 
between researchers and developers. Specific communications objectives 
of the SEL are to: 

Devise software development standards 

Develop software management guidelines 

Provi Je real-time teedoack to development teains being monitored 

Maintain contact with the software engineering research community 

Clearly, the objectives of the SEL reflect its multistep approach to 
software enyineeriny, as described earlier. 

25 2 




The general cateyory ot fliynt dynamics software includes applications to 
support attitude determination, attitude control, maneuver planning, 
orbit adjustment, and yeneral mission analysis. Most of these proyrams 
are scientific and mathematical in nature. The attitude systems, in 
particular, are a larye and nomoyeneous yroup of software that has been 
studied extensively. The attitude determination and control systems are 
designed similarity for each mission using a standard executive support 
packaye as the controlling system. 

Typically, attitude systems read sensor measurements from a telemetry 
stream and determine the attitude of the spacecraft from this data. 
Oependiny on cne types of the data available and the accuracies required, 
the size of these systems may range from 3O,.OU0 lines of cjde to about 
120, UoO lines of code. All these systems are desiyned to run in batch 
?nd/or interactive graphic mode. Some existing software can be reused 
for each new system, since there are always some similarities to past 
systems, especially in the high-level design. The percentage of reused 
code ranges from lo percent to in upper limit of nearly 70 percent. 

All applications developed in the flight dynamics area of G.SFC have 
developmeiit time constraints corresponding to launch dates. Most of the 
software discussed in tnis paper must be completed (implying completion 
of acceptance testing) 6l) days before the scheduled launch. Il the 
software is not completed, required capabilitias must be deleted or 
redetined, and an alternative versior, of the intended system must be 
defined to ensure that the mission can be supported in some limited 

The development process normally begins approximately 16 to 24 months 
before a scheduled launcn in order to be completed 2 montns in advance of 
launch. This developinenc period is divided into phases typical of i"hc; 
standard software life cycle. 

'1 25 ^ 



+j^. ^ '■■.•■■iV'' 



Tht data provided by the SEL has formed the basis ot numerous software 
enyineeriny studies. The software development tasks from which data was 
collected for the SEL data base are comprised of over 5U fliyht dynamics 
projects developed over 8 years. All data collected by the SEL is 
assembled in a computer data base to facilitate its access by researchers 
and manayers. 

Two very strony effects were identified early in the SEL Investiyations 
and have been confirmed in the literature (Reference ]] . That is, 
variation in programmer abilities appears to be the most powerful 
influence on the productivity and quality of software development. In 
adoition, the nature of the local computmy and work environments seems 
to be a iiqnificant determining force on the process. Any valid 
experimental desiyr, which is attempting to study effects of 
parameters such as methodology must account for or eliminace these 


lAiiMMt -^'X*,. 



A software development methodoloyy is the reyular application of a set of 
specified techniques to part or all of the software development process. 
The methodoioyies and techniques studied Dy the SEL can be classified 
into five yroups. The yroups and some exanples of edch are listed below: 

Oesiyn Techniques 

- Top-down structured desiyn 

- Tree charts 

- Data flow diayrams 

- HIPO charts 

- Process desiyn anyuayes 

Desiyn Evaluation Techniques 

- Strepyth and coupliny analysis 

- Connection matrices 

- Proyrari; correctMess proofs 

Structured Implementation Techniques 

- Top-down structured prouranininy 

- Structured 1 anyuayes 

- Code readiny 

- Walkthrouyhs 

Manayement Tecnniques 

- Chief proyrarmier teams 

- Oesiyn reviews 

- Librarian functions 
Independent lest teams 

[locunentation Techniques 

- Automated documentdtion syst^:ms 

- SLT'-ictured code 

The SLL's approach to evdlu?tiny methodoloyies nas oeen to co'' ■ ,'. 
add quality data from similar p. ejects triat employed differ^'' 
dev?Iop(nerit irethodoloyies ( semicof •to ! l^d) experiments). Tin rye 
effects of the methodoloyies on the product can then be >- and the 
useful techniques identified. Controlled experiments would be thv-> ideal 
iit-ans of collectiny data for these anciiyses. However, the cosr of 
duplicatmy any larye development effort precludes that straLcjy. 

The inibility to make complete comparisons of the projects studied has 
delayed the derivation of definitive conclusions from the data. However, 
some effects are apparent A summary of the early results of methodoloyy 
evaluations is presented in Table 1, A superficial examination of tnis 
table suyyests the reasonable cor.clusion thaL most techniques that do not 


K^ --r.> 



siyni f icantly increase the proyrammer's and/or designer's workload Du^ 
tnat provide a hi;jher level of oryanization to his/her activities have a 
positive impact on the development process. 



fc'.-^kicWIki *C%-v •» 



Cost Effective 

Formal Test Plan 

Process Uesivjn 
Ldnyu:iye ^PDL) 

Coae Readny 

Formal Traininy 


Structured Analysis 
Conf lyuration Mynit. 
Desiyn Formalisms 

Fonnal Desiyn Reviews 

Structured Code 
(Precomp^ lers) 

Iterative Refinement 

Results of Evaluations 

Cost Unclear Not Cost Effective 
Code Walkthrouyhs Simulated Constructs 
Top-down Desiyn Axiomatic Desiyn 

Top-down Code 

Chief Proyrammer 

Code Auditors 


Automated PUL 

Cocie Analyzers 

Larye Problem Statement 

Independent Verification and 
Inteyr ation 

Reliability Models 

Automated Flow- 
charf -^rs 

Resource F^timation 




. *• <-,•*' v*V V • i^-' 

Mcce rigorous techniques have been applied to the analysis of some 
subsets of the SEL data on methodoloyies. Table 2 shows the results of 
study of the effects of methodology on productivity. Essentially, it 
confirms the SEL's earlier conclusions. 




Formal Oesiyn Review 

L-'sign Formalism 

".-aiijn Decision Notes 

'_ .-sign Walkthrough 

Code Walkthrough 

Cocie Reading 

Top-Down Design 

Structured Ci. Je 

Liorarian Use 

Chief Programmer Team 

Fonndl Test Plans 

Heavy Management Involvement 

Formal Training 

Top-Down Code 


























- '--»f.-~V- 

^ iS.^iI'i4.,C:, ^*'_. ,'..-.. " • U 

In addition to experimenting with and measuriny the above mentioned 
technologies, the SEL has gained insight into additional areas which may 
have major impacts on the overall productinty as well as quality of 
software development. 

Productivity for large vs. small systems 

The common belief by many software managers and developers is that as the 
size of a software system increases, its complexity increases at a higher 
rate than the lines o^ code increase. Because of this fact, it is 
commonly believed that in the effort equation 

E = alb 

where E = effort of person time 
where I = 1 ines of code 

that the value of b must be greater than 1. The projects that the SEL 
has studied have been unable to verify this belief and instead have found 
the value of b to approximately .92 in the SEL environment. The fact 
that this equation is nearly linear leads to the counter intuitive point 
tnat a project of IbO.OUO lines of code will cost approximately 3 times 
as much as a 50,000 lines of code project instead of 4 or b times as much 
as is often commonly believed. 

(Further details can be found in Reference 2.) 

Productivity Variation 

Another characteristic that the SEL has been interested in studying has 
been the variations in programmer productivity. Obviously one would want 
to increase the productivity oy whatever approach found to be effective, 
but first we must clearly understand what the baseline characteristics 
of productivity aie (minimum, maximum, average, difference between small 
and large projects, etc.); oniy then will we know if we have improved or 
not in the years to come. 

As has been found oy otner researchers in varyiny environments, the 
productivity of different progranimers can easily differ by a factor of 8 
or iL to I. Tne SEL did Tind tnat there was a greater variation (from 
very iow productivity of .5 l.o.c./hour to 10.8 l.o.c./hour in small 
projects. The probable reason tor this is that newer people are 
typically put on smaller projects and the SEL has found extreme 
d'tferehces in the relatively inexperienced personnel. 

R ^-using Code 

As was stated in tl'e introduction, projects being developed in the SEL 
environment typically utilize approximately 30 percent old code. 
Although it is obviously less costly to integrate existing code into a 
system rather than having to generate new code, there is some cost that 
must fie attributed to adoptiny the old code. The development team must 
test, integrate and possibly document the old code, so there is some 


*■ '"^■.**^ 

±1.0^^''%^-, •'•• ■ {± 

overhead. By lookiny at approximately 25 projects ranging in size from 
25,000 lines of code to over 100,000 total lines of code and ranying in 
percent of reused code from percent to 70 percent, the SEL finds that 
by attributing a value of approximately 20 percent overhead cost to reuse 
code the expenditures of the 25 projects can best be characterized. Now 
the SEL uses the 20 percent figure for estimating the cost of adopting 
existing code to a new software project. 

Development Resources 

Another area of concern to the SEL in defining the basic profile of 
software development, was that of staffing level and resource expenditure 
profiles. Many authorities subscribe to the point that there is an 
optimal staffing level profile which should be followed for all software 
projects. Such profiles as a Rayleigh Curve are suggested as optimal. 
Chart 8 depicts characteristics of classes of projects monitored in the 
SEL and shows the difference in productivity and reliaoility for groups 
of projects having different staffing level profiles. Although the 
Rayleigh Curve may be acceptable for some projects, the SEL has found 
that wide variations on these characteristics still lead to a successful 
projects. The SSL has also iound that, extreme deviations may be 
indicative of problem software. 

Resources Allocation 

One set of basic information that one may want to understand is just 
where do programmers spend their time. When the SEL looked at numerous 
projects to understand where the time was spent, it found that the SEL 
environment deviated somewhat from the old 40-20-40 rule. Typically 
projects indicated that when the total hours expended were based on phase 
dates of a project (i.e., a specific data defined the absolute completion 
of one phase of the cycle and the beginning of the next phase) the 
D-eakdown was less than 26 percent for design, close to 50 percent for 
Code and about 30 percent for integration and test. 

When the programmers provided weekly data attributing their time to the 
activity that they felt they were actually doing, no matter what phase of 
software development they were in; the profile looks quite different. 
The three phases (designs, code, test) each consumed approximately the 
same percent effort and over 25 percent of the time was attributed to 
'other' activities (such as travel, training, unknown, etc.). The SEL 
has continually found that this effort ( other ) exists, and cannot 
easily be reduced, and most probably should be accepted as a given. The 
SEL has found it to be a mistake to attempt to increase productivity 
merely be eliminating iiiajor portions of this 'other' time. 

Cost Models 

In addition to the studies made pertaining to various measures fo'- 
software, the SEL has also utilized the cost data collected from the many 
projects to calibrate and evaluate various available resource estimation 
models. No attempt was intended to qualify one model as being any better 
than another. The objective of the studies was to better understand the 


sensitivities of the various models and to determine which models seemed 
to characterize the SEL software development environment most 

In studying these resource models, nine projects which were somewhat 
similar in size were used as experimental projects. Each of the models 
was fed complete and accurate data from the SEL data base and each 
was calibrated with nominal sets of projects as completely as tha 
experimenters could. Summary results indicate that, occasionally, some 
models can accurately predict effort required for a software project. 
The SEL has reitterated what many other software develops and managers 
claim. Cost models snould never be used as a sole source oi estimation. 
The user must have access to experienced personnel for estimatiny and 
must also have access to a corporate memory which can be used to 
calibrate and reinforce someone's estimate of cost. Resource irodels can 
be used as a supplemental tool to reinforce one's estimate or to flay 
possible inconsistencies. 

More detailed information on the SEL studies can be found in Reference 3, 
4, and b. 

Effects of MPP on Software Development 

In an attempt to determine if the utilization of Modern Proyramming 
Practices (MPP) has any impact (either favorable or unfavorable) on the 
development software, a set of 10 fairly large (between 50,000 l.o.c. and 
120,000 l.o.c. and fairly similar projects (same developtnent environment, 
same type of requirements, same time constraints) was closely examined. 
These projects had been developed in the SEL environment where detailed 
information was extracted from the projects weekly and where each project 
had a different level of MPP enforced during the development process. 

The MPP's ranged from various design approaches (such as POL, Design 
Walkthroughs, etc.) to coda and test methodologies (such as structured 
code, reading, etc.), to various integration and syster testing 
approaches. All the possible MPP's were rated and scaled as to the level 
to which the practice was followed for each project (the rating was done 
by the SEL researchers Dy the software developers). The only purpose of 
this exercise was to depict trends and not to prove that any one single 
practice was more effective by itself than any other. 

The level to which MPP's were utilized were plotted against productivity 
against error rate. The application of the MPP has favorably affected 
productivity by about lb percent for these experiments. Kesults of 
software reliaoility vs. MPP is very questionable. The SEL is continuing 
analysis of additional data. 

(Kore details of this effort can oe found in Reference 6). 




■- r 

Co nclusions 

Several points stand out amony the results of the research performed in 
the SEL. These tv;o points are as follows: 

1 . The software development pro c ess can be improved through the 
applic ation of selected methodol gies. This >jeneral conclusion wa s 
derived from observations made dunny the past several years. 
Productivity rates have steadily increased throuyh the years with the 
application of more refined methodologies. Even with the additional 
overhead of data collection and special training, a steady improvement in 
the development process is evident. 

The amount of improvement attributable to any given methodoloyy is yary 
difficult to quantify, but the history of the SEL indicates that almost 
any of the disciplined mathodologies available will favorably affect the 
process by about 5 to 10 percent over the absence of any such approach. 
A methodology that is particularly well suited to a specific environment 
coi ^d enhance productivity by as much as 20 percent. Optimizing the 
organizational structure of the people supporting the project can produce 
an additional improvement of 10 percent. 

2. The greatest need is for the rational application of the available 
tecnnologies, not tor the creation of new technologies. During the past 
several years, the SEL has learned that there are no shortages of well- 
defined methodologies and tools. The deficiency of current, practice is 
in the utilization of the available software technology. Software 
implementers have Deen slow to evaluate and adapt these approaches to 
their particular environments. 

Software technologies snould not be accepted without critically examining 
their effects and without understanding the environment in which they 
operate. However, the evidence is conclusive that the software 
development process can be substantially improved throuyh the application 
of appropriate t'.c^noloyy. 




i .^M^vU 'I 



1. B. A. Shell, "The Psychological Study of Proyramming," Computing 
Surveys. March 1981, vol. 13, no. 1, pp. 101-120 

2. V. R. Basil i and K. Freburger, 'Programming Measurement and 
Estimation of the Software Engineering Laboratory', Journal or Systems 
and Software , February 1981, Volume 2, No. 1 

3. Software Engineering Laboratory, SEL 81-104, The Software Engineering 
Laboratory , 0. N. Card, F. E. HcGarry, G. Page, et. al ., February 1982 

4. SEL 80-007, An Appraisal of Selected Cost/Resource Estimation Models 
for Software Systems, J. F. Cook, F. E. McGarry, December 1980 

5. V. R. Basili, 'Software Engineering Laboratory Relationships for 
Proyrainning Measurement and Estimation', University of Maryland, 
Technical Memorandum, October 1979 

6. SEL 82-001, Evaluation of Management Measures of Software 
Development , 0. Card, G. Page, F. McGarry, September 1982 


■•. J^^if*^". 


1 ^f^M^rJ, ■ \± 






Phillip P. Marino 
Bendlx Field Engineering Corporation 


Nerf methods and approaches are being tried and evaluated with the 
goal of increasing productivity and quality. The underlying concept in 
all of these approaches, methods or processes Is that people require 
Interactive communication to maximize the organization's strengths and 
minimize impediments to productivity improvement. 

This paper examines Bendlx Field Engineering Corporation's 
organizational structure and experiences with employee involvement 
programs. The paper focuses on methods Bendlx developed and Implemented 
to open lines o" communication throughout the organization. The Bendlx 
approach to productivity and quality enhancement shows that Interactive 
communication is critical to the successful Implementaiton of any 
productivity improvement program. The paper concludes with an 
examination of the Bendlx methodologies which can be adopted by any 
corporation in any industry. 




4 s. •S'''?«'aw V' 


Bendix Field Engineering Corporation's (BFEC's) international 
reputation In the technical field has remained an industry standard for 
more than 35 years. Quality service, combined with effective 
competition in the world market, has contributed to BFEC's distinction 
as a preeminent supplier of engineering and support services in the 
electronics and aerospace industry. 

In a field as dynamic as the technical service industry, with a 
multlorganlzational structure, BFEC has maintained its position by 
unremitting efforts to enhance quality, productivity and cost 
effectiveness of the services provided to its many customers. 

BFEC has long had a 'working smarter' concept which reflected the 
corporate commitment to engineering excellence and productivity 
improvement. Many factors have contributed to productivity and quality 
enhancement, but the most crucial and Indispensable "element Is the 
quality of service provided by our skilled engineering and technical 
professionals supported by oi>r progressive management team and dedicated 
work force. 

The purpose of this paper is to present methodologies BFEC 
created and adopted to augment a successful productivity and quality 
enhancement program. 

BFEC's Productivity and Quality Enhancement Program did not start 
]ast year. The foundations of the program have been in place for 
years. These foundations Include statistical concepts and 
accountability, goal setting, and measurement of performance against 

In the labor-intensive service Industry, for example, 
productivity can be directly related to absenteeism. BFEC has 
con.>^-tar.tly searched for ways to reduce absenteeism so that productivity 
may be increased. Records of absence rate and associated costs have 
been maintained since 1960. In 1976, an alarming trend of increased 
absence was noted. Armed with a long history of absence statistics, the 
Absenteeism Reduction Program was developed. The program begins with 
management reports of absence trends by individual departments, and 
includes research information, company determined standards of 
performance, recognition of employees with outstanding attendance 
records, rewards, publicity of the program, and goals for each element, 
department, directorate and the corporation as a whole. As a result of 
the implementation of the Absenteeism Reduction Program, absence was 
reduced from 2 52% in FY 76 to a rate of 1.91% for CY 84 (refer to 
Figure 1). BFEC has avoided spending over $3.2 million dollars in lost 
wages alone during the past 9 years. 



Mtov Kin...- 




JOST S-1 ,iN6S 


FY 77 
FY 79 










1 80 


?10 220 




2 SO S3,20S.Sn 



Another example of the continuous emphasis on productivity and 
quality enhancement Is the Cost Reduction and Suggestion System. 
Established in 1963, the program has motivated employees to seek better 
ways to perform their jobs while maintaining or Improving the quality 
and reducing the costs associated with performing their jobs. Figure 2 
depicts the cost savings of more than $18 million since 1981. 







An illtistration of the corporate commitment to quality, 
productivity and cost effectiveness is demonstrated by one of our 
contracts with NASA (figure 3). Originally implemented in 1978, this 
contract began with a staff of 420 employees. After the contract was 
awarded, it was converted to BFEC's first Mission Contract. On contract 
renewal in 1983, our tasks had been expanded to include maintenance of 
approximately 50 computer systems and peripherals in addition to control 
center maintenance and operations duties at the Goddard Space Flight 
Center. BFEC's attention to zero-based staffing analysis, cross 



- •'*-i,»*^-- 




IM*'' lANCF 








training, and cross-utliizatlon of personnel, plus streamlining of 
operations allowed the new contract to begin with a staff of 84 fewer 
employees than the original contract. All of this was accomplished 
through transfers and normal attrition and exclusive of all NASA mission 
precipitated staffing changes. Personnel reduction also resulted in 
lower costs while, at the same time, keeping maintenance and operation 
proficiency at above average levtls. 

"We believe that each generation of products and tech- 
nology should surpass the excellence, ingenuity and per- 
formance of its predecessor and that all our products 
and service must reflect the integrity that is one of 
our most precious assets." 

These -vords, from the Beiidix Creed written over 40 years ago, 
challenge eai^h employee to practlc" productivity inprovement and quality 
enhancement as a way of life. The employees are constantly challenged 
to learn something new every day. BFEC's President remarkti that "a day 
when you don't learn a better way of doing your job is wasted day". 
With these types of challenges, BFEC took steps to consolidate and 
formalize the Productivity and Quality Enhancement Program tn 1984. 

The first step was taken by the President with the establishment 
of the Productivity Council, a group of seven individuals from various 
major departments who were committed to the basic tenet that a 
formalized quality and productivity improvement program is essential for 
BFEC to remain a leader in providing engineering and support services. 
The initial group became the product champions of the productivity cause 
within their respective organizational units. 

They did not charge ahead without understanding BFEC's past 
performance — both achievements and failures. The council studied and 
learned from corporate experiences. 




¥i ^it!<?k.\.i:. "^ : ■ (-^ 

The coui"»cll was aware of a major Bendlx Corporation campaign for 
productivity Improvement in 1980. During that campaign, the 
Productivity Committee did research, provlr'ed training to management In 
national ar.d International trends In produs tivlty, developed a catchy 
theme and poster design, and conducted a siv^rt term productivity 
suggestion program. This short term program resulted in over $500,000 
in cost reductions in a three month period, but the program was not 
sustained because there was no continuing long term commitment and 
because a consolidated productivity improvement and quali.y enhancement 
program \:?.d not been established. 

In 1978, Bendlx introduced an objectives-based performance 
planning and review system (PPRS). This program included over 16 hours 
of training for managers In the f. w or process Involved in planning and 
measuring performance. The PP&RS, as it is now known, is used 
throughout the company, but initially there was some reluctance to fully 
Implement the philosophy of the program. Rather, there was an attempt 
to fill out the associated forms, and not practice the management 
philosophy it espoused. 

In an effort to be more productive, the I.'&RS was the subject of 
reevaluation in 1982. As a part that study, a special task force asked 
for and received Information from employees, supervisors and managers 
about their refictions to the system. The task for found that, 
although the program did Improve communication, documentation for the 
system was limiting and the task force suggested a more free-form 
format. As a result of the study, employees began to implement the 
philosophy of the program more effeccively. 

The council learned that major programs cannot be Implemented 
fully on a dictatorial basis. It also learned that when ma .agers, 
supervisors and employees have an oppoi'tunlty to participate in the 
structuring of a program, and have an active role In the development of 
a program, there a greater acceptance — even owrn rship of the concept. 

With an awareness of FFEC's accomplishments, an awareness of 
programs which were deemed to be less than successful, and a realization 
of the basic employee need to be part of a developing program, the 
Productivity Council began the task of formalizing the Productivity 
Improvement and Quality Enhancement Program for BFEC. 

The council developed a logo 
for the emerging program. The logo 
is baacially a star, the symbol of 
achievement and excellence. But 
more than just a star, it is people 
joined together to represent the 
entire employee population working 
together as a team. Neither a star 
nor people joined together as a 
team are new concepts; but the 
final product — this logo — ii; 


w k^ . ^. ■<>,•;». v*i 



The next step in the formalization of the BFEC Productivity 
Program was to discuss the long standing productivity and quality 
commitment with all levels of employees, from executives to entry level 
employees. The Productivity Council compiled the statement of BFEC 
Productivity philosophy with the following primary components: 

o Improving communications and cooperation vertically and 

horizontally within the corporate hierarchy and adopting a 
more participative style of management 

o Educating all management persor 2I and employees In 

techr'ques designed to identify, evaluate, and solve problems 

o Continuing the quality improvement process aimed at "doing 
it right the first time" for tasks performed within all 

o Using statistical techniques throughout the company: 

operational statistics, cost control, manageuient , absence, 
safety, etc. 

o Expanding opportunities for employee participation in the 
Improvement of services supplied to BFEC's customer 

o Encouraging supplier or subcontractor participation in the 
quality/productivity improvement process and providing 
assistance and direction, as required 

The Productivity Council recognized that an effective program 
required acclve participation from all employees at all levels of the 
organization. The President and two key Vice Presidents were chartered 
with the responsibility for the final administrative and budgetary 
decisions as the Productivity Program's Executive Committee. The 
Productivity Council would continue to consist of a Director and 
representatives from key operational and administrative organizations. 
Individuals from each department volunteered "0 serve as Department 
Productivity Coordinators. Employees were surveyed and given an 
opportunity to participate in BFEC's adaptation of the productivity 
improvement process. The council evolved a significant and unique role 
for a vital gr'^up of employees — the managers and supervisors. 

The council quickly recognized that one potentially effective 
approach to augment organizational appeared to be a merger 
of quality assurance methodology and behavioral science theory. The 
council researched and examined existing literature on the subject and 
contacted companies that had implemented or were in the process of 
implementing productivity improvement programs The council developed 
the Productivity Enhancement Team (PET) concept, and tried the concept 
out In one department as a prototype. This department was chosen 
because it had a long record of productivity and quality commitment 
demonstrated by marked open communications, an effective team of 
managers, and a desire to be innovative. 

4 270 

• - • '#**!^. 



Research revealed that most existing productivity programs 
centered around manufacturing concerns. The measurement of success was 
typically based on productivity or quality enhancements to a product or 
consumer Item that could be easily measured. However, research into 
motivation and employe? Involvement principles indicate that these 
manufacturing-oriented productivity programs could be transferred with 
some customization to the service Industry. Through an iterative 
process, traditionally structured productivity improvement methods were 
tailored to meet the needs of the BFEC organization. 

The growth of the PET program has been controlled during its 
evoluMonary phase. Each PET is made up of volunteers, from four to ten 
members, who meet on a regular basis to discuss productivity 
Improvement, problems, concerns, and solutions. Employees have been 
enthusiastic about joining and working in these employee-run teams. 
They work on problems appropriate to their job responsibilities. 

Success of PET teams Is linked to certain key elements. The 
voluntary nature of the PET is an Important ingredient in developing 
employee support, commitment, and enthusiasm. The voluntary nature of 
the program is rooted in the basic concepts of motivation and employee 
Involvement. PET members learn to work as a team, which not only 
encourages problem solving but also encourages joint effort and 
collaborative ways of working. This 'team effort' is a new experience 
for many employees who typically perform tasks alone or in a limited 
exposure to others. 

The most significant aspect of the PET process is the structure 
in the program. Many employees have little experience and few 
opportunities to participate in effective structured problem analysis or 
to work successfully in group settings. During BFEC's PET training, 
employees undergo six weeks of orientation to widely accepted 
productivity Improvement methodologies Including: problem definition; 
brainstorming, data gathering; cause-and-effect analysis; Pareto 
analysis; preparation of data check sheets, graphs, and charts; and 
presentation techniques. The purpose of this training is to ensure that 
all PET's operate within the same guidelines, using a specific approach 
to problem solving with the goal of presenting a solution to a work 
related problem to supervisors and managers in a method which easily 
demonstrates that the problem and recommended solution were thoroughly 

As BFEC implemented the PET process in a controlled environment, 
it became apparent that supervisors dnd managers had a very important 
role in this evolutionary program. The supervisors of the first PETs 
had mixed reactions. Some were apprehensive because the PET identified 
problems which the supervisor thought should have been recognized and 
solved by the supervisor. Some supervisors were fearful that the PET 
would usurp their authority. 

The Productivity Council discovered from Its research and 
practical experience that a productivity program such as the one 
described cannot succeed without Involvement from management. To 
achieve this lnvol\'emont and support often requires a drastic change in 
traditional management attitudes. 






True change cannot be dictated or mandatsd. It must be truly 
embraced by the highest level of management and communicated throughout 
the company. Communication and training starts with the company 
president and proceeds to directors, program managers and senior 
managers, department managers, and supervisors. 

The council contacted various vendors of management development 
and training and vendors who specialized in productivity programs and 
quality of worklife programs. They asked for ways to develop meaningful 
roles of supervisory and management personnel. 

The vendors submitted proposals for 4 to 16 hours of tralniag, 
during which the supervisors and managers would experience the PET 
process and discuss their reactions to the group process. The council 
felt that somethlnc was missing. 

They looked and found a resource who talked about the process of 
change. They realized that to achieve true system-wide acceptance of 
the PET process, managers and supervisors would have to not only support 
and endorse the structured PET techniques, but demonstrate these 
techniques in their day-to-day interaction with employees. 

Given BFEC's long standing pattern of productivity improvements 
and proven methods of cost efficiency, we onJy had to provide general 
guidelines for supervisors and managers. The Productivity Council 
compiled the Productivity Procedures Manual which incorporates existing 
programs, procedures, reporting channels with the PET techniques, 
recognitions and rewards. 

Let's review the steps taken by the Productivity Council. They 
established the PET program, a structured process for small groups of 
employees to discuss work related problems, research these problems and 
present well researched and documented solutions to management. 
Throughout the PET process, supervisors and managers maint n close 
contact with the PET through receipt of the minutes of each meeting, by 
responding to PET member requests for technical information, data, and 
support, by participating, as guests, at regular PET meetings and by 
evaluating and responding to PET recommendations. 

Realizing that department management may not be able to implement 
a PET solution due to budget or manpower constraints, the Productivity 
Council further opened a new line of communication between employees and 
the highest level of management by providing the PET with the 
opportunity to make its presentation directly to the council. The 
council then has the responsibility to approach the Executive Committee 
which has the ultimate power to provide additional budget, manpower and 

The Productivity Council has further developed a forum to open 
lines of communication between supervisory and managerial employees by 
requiring an annual departmental productivity plan. This plan is not 
designed to respond to the needs of the Productivity Council. It is 
designed to cause the department director, managers and supervisors to 
meet together and, using PET-type techniques (brainstorming. 




problem-solving, etc.). develop specific measurable productivity and 
quality goals and objectives for the year. The Council does not expect, 
nor does It want, a glossy productivity plan which Is outside the 
current and established work of a department. It seeks the 
documentation of goals and objectives which are part of the way the 
department does business. 

In the formalization process, the council opened lines of 
communication between employees by providing the structured productivity 
improvement techniques. It is important to examine the lines of 
communication and the formal and Informal channels for communication. 
From this examination, BFEC can share with other service groups its 
experiences in re-emphasizing productivity as a way of life within an 
organization. BFEC has opened informal lines of communication between 
employees and managers by providing an opportunity for employees to 
study work problems and recommend solutions to management; opened lines 
of communication between executives and managers and supervisors by 
providing a forum for the determination of departmental productivity and 
quality goals and objectives; and opened a new formal line of 
communication between employees and the highest level of management by 
providing a PET with the opportunity to make a management presentation 
directly to the Productivity Council. 

The BFEC Productivity Council has tried to Involve all employees, 
at all levels of the various organizations, in the productivity 
Improvement process. All of our approaches, methods and processes 
require people and people require Interactive communications to be 
optimally effective. 

Together, management and employees must communicate tc maximize 
the organization's strengths and minimize impediments to oroductlvity. 
I People need to know their role in accomplishing the company's goals and 

objectives and they must be able to communicate to management their 
desire to accomplish these goals. 

People are the strength of the organization and, therefore, the 
mainstay of the productivity and quality effort. They help the 
organization use its resources more efficiently and effectively for the 
betterment of the company and its customers. 

What has happened in the last year and a half is systematic 
formalization of the process throughout the organization. Components of 
the BFEC Productivity Improvement and Quality Enhancement Program 
Include research, reviewing past achievements and failure, learning from 
other companies and from our own experiences, adapting structured 
methodologies of effective productivity improvement programs from 
tj.ically manufacturing entitles to the fluid, responsive, mobile world 
of the service industry, and formulating a program which involves 
personnel at all levels of the organization. 

The ideas and methods in this paper are not new. They are 
distinctive. They are like the BFEC productivity logo — a star 
composed of people interacting. Individually the pieces are well known 



^^„>^t%v-, ^, ■ ■ • " (VJ 

but together distinctive and a symbol of a renewed vigor to work 
together, to work more Innovatlvely, to work smarter. 

(Phillip P. Marino, Senior Manager of Software and Engineering Services, 
began his career with the Bendlx Field Engineering Corporation as the 
Software Quality Assurance Manager. He is now responsible for all 
software and systems engineering services on the Mission Operations 
Support Services Program for NASA real-time spacecraft control 
netuorks. His prior corporate experiences Include Corporate Systems 
Manager for the largest Import automotive distributorship on the east 
cost, responsible for all ADP activities for the corporation. Further 
experiences include over 16 years in data processing with the Maryland 
Department of Transportation, Department of Health, and Department of 
Education, culminating as Deputy Director for the Department of 
Education. Mr. Marino holds an Executive MBA degree from Loyola College 
of Baltimore, Maryland.) 







George L. Carter 
Manager, Socio-technical Engineering 

Westinghouse Manufacturing Systems & Technology Center 
Columbia, Maryland 21046 


The objective of socio-technical theory and design 
is to provide the best match between the social system and 
the technical system. The achievement of a best match 
makes optimal use of the resources of both systems. 

Implementation of this theory is best served when 
there is involvement by the "user" organization. The 
"involvement" relative to the introduction of new 
technology in the organization is extremely significant. 
Employee involvement is critical to effective participative 
management. Because this style is considered to be "new", 
many managers lack the experience in dealing with such an 
approach to management. 

The trends toward participative management and 
employee involvement have taken various forms. These have 
included quality circles, semi-autonomus teams and adhoc 
action teams. It is noteworthy to point out, as these 
processes have evolved, the role of a facilitator has 
become more prevelant. 

The facilitation of the socio-technical design 
system will use the t-^ols of indurtrial engineering and 
will be managed usinc a style which has been structured 
under behavioral science principles. 

The successful integration of new technologies into 
the business organization today will be predicated on the 
socio-technical system which has been developed. 



J-., v*r'i^kvl-. ^ 


Socio-technical theory originati.ig in the 1950s 
looked at work effort as a system of technical and social 

"Socio-technical design is based on principles from 
two different worlds between which a fundamental 
split exists. The technical world is organized 
around rational principles of efficiency, while the 
phenomenal world within which humans live their 
daily lives is organized around psychological 
principles based on cognition and emotion. The 
principles by which one world is organized are not 
necessarily or even likely to be the same as those 
by which the other world is organized. 

The objective of socio-technical design is to bridge 
the two worlds through a "best match" between a 
social system organized around phenomenal-world 
principles and a technical system organized aroun*^ 
technical-world principles. The achievement of a 
best match makes optimal use of the resources of 
both systems . The resources of the two systems are 
optimally used when the two systems are members of 
the socio-technical system. 

A work group is self-regulating to the degree that 
group members define themselves as system 
contributors rather than job-holders." (Susman 1976) 

Perfornence, relative to this theory, is most 
effective when the social and technical components are well 
matched to each other. This match does not just happen — it 
must be properly planned and implemented. This 
implementation is "best" served when there is involvement 
by the "user" community or organization. 

The involvement of the "user" relative to the 
introduction of new technology in the organization is 
extremely significant. For decades the organization of the 
workplace had been predicated on the principles of 
icientific management. This system is still routed in the 
drive for more output per hour, or breaking down each job 
into its simplest, most repetitive specialized tasks, with 
the shortest possible learning period. Scientific 
management had concentrated on the technical exclusion of 
the social. 


■■• ^tf^-^. 


This management philosophy generally considered that 
the worker involvement was of limited value in the 
organization. This is no longer true. 

Employee involvement is critical to effective 
participative management. Because this style is considered 
to be "new", many managers lack the experience and know-how 
in dealing with such an approach to management. Sometimes 
the concept is viewed by managers and supervisors as a 
threat in terms of conventional power and authority. The 
impatience of some managers to achieve the short-term 
economic gains while dealing with a sensitive new product 
that requires long term commitments forecasts, at best, an 
uneasy pathway to meaningful results. 

Me can anticipate, therefore, that the worker 
participation in the development of these kinds of new 
programs to improve the so-called "quality of work life" 
will manifest itself in a step-by-step effort to enhance 
the welfare of the workers and to uplift their human 
dignity. For the present and forseeable future, worker 
involvement in decision-making will more readily spring up 
with regard to the more immediate and noticeable aspects of 
the working life. Managing the job is more immediate and 
urgent. The workers' concern for the managing of the 
enterprise is best measured by the immediacy of the impact 
on the worker himself. 

The trends towards participate management and 
employee involvement over the last several years have taken 
various organizational forms. These have included quality 
circles, semi-autonomous teams and ad hoc action teams. 
Participants in these organizations for the most part have 
been voluntary. It is noteworthy to point out, however, as 
these processes have evolved, the role of a "facilitator" 
or "third party consultant" has become more prevalent. 
One can logically conclude that as participative management 
philosophies grow, the role of the facilitator and third 
party consultant will, likewise, grow. 

It should be noted, however, that the role of the 
facilitator should, in time, become a management style and 
philosophy and not a position identified with a "program". 
"Programs" traditionally become the "flavor of the time" 
and succumb to new programs that take their place. 




■■?^.:Ar O'! 


In one organization in which I was associated, a 
large quality circle activity prevailed. The quality 
circle facilitators of the organization each had in excess 
of twelve circles in progress at any given point in time. 
The organization had a need for more circles, but was 
hesitant about adding additional new facilitators. To 
address the problem, "ownership transfer" was instituted. 

The process called for the quality circle team 
leaders which were work section supervisors, to be trained 
to become facilitators, and secondly, a member of the 
quality circle to become the team leader. The original 
facilitator became a member of the Operation Manager's 
staff and functioned as an advisor and consultant to the 

This "ownership process" accomplished two purposes: 

1. Allowed the establishment of additional 
quality circles without the hiring of more 

2. Provided the first line supervisor with a 
"facilitation" management style which 
helped to develop a participative 
management environment in the organization. 

The effectiveness of the circles to enhance the 
organization became significant when the objectives of the 
organization were shared with the "circles", and the 
subsequent projects of the circles dealt specifically with 
supporting and accomplishing the objectives, thus providing 
a focused activity for the circles. 

A second organization which I was familiar, was 
organized specifically to utilize the participative 
manaoement concept. The organization was structured around 
semi-autotomous teams. Instead of the traditional 
first-level supervisor, the teams had team leaders or 
facilitators. The facilitators provided the necessary 
training and technical support to the teams so they could 
accomplish the established goals and objectives of the 
organization. The roles of the "team" members changed 
regularly from "production workers" to support functions 
such as production and quality control. 

The team leaders of the organization had 
professional classifications and had come from industrial 
engineering, production, quality, and test backgrounds. 



• ".'■jfrsf-V-. 

1 f 

The effectiveness of these organizations has been, 
for the most part, dependent on the training given to the 
participants. The training provides analytical and problem 
solving skills used traditionally by industrial engineers. 

The industrial engineer played a primary role in the 

scientific management of the organization, both from 

administration of the process and the controlling 
functional responsibility. 

With the advent of the socio- technical process, the 
:^ industrial engineer will again play a major and significant 

' role. This basic logic is credible, however, there is one 

basic flaw. 

' Traditionally, the industrial engineer lacks 

training in the behavioral sciences which is essential in 
effectively interacting as the "third party consultant". 
The role significantly changes from the traditional "tell 
them how" to the "suggestive" approach as a technical 
resource to the participative group. 

'\ For the present time and immediate future, 

specialized training in the behavioral science principles 
and application must be provided to the industrial 

- engineers. 

\ This aspect of education is not presently required 

in most t.i-aditional Industrial Engineering curriculums. 

Success of the socio-technical theory is predicated 

on a participants environment which involves the commitment 

I of the "user". The greater -he amount of user awareness 

and commitment to the process, the lesser amount of 
resistance to change and the new technology. 


r It is very difficult to develop a participate 

approach in an organization when the members have been 

|_ operating independently of another. Now, to use their 

] individual personalities, work habits, and ideas toward a 

^ group goal is a very ambitious undertaking. There's no 

1 denying the beneficial impact that genuine participation 

\ can produce and the positive repercussions it will have in 

, other areas. 


•• - ■^c-Sa-. 

■4 K^-r':jC<' 


In adopting a participative concept philosophy, the 

most often asked question is, "What is in it for the 

organization's menibers--and what benefits can the 

organization obtain? We will first take a look at the 

research that supports why organization members want to 

The often-used cliche that we hear, in justifying a 
participative concept today, is that "people tend to 
support what they help create." This statement is perhaps 
becoming more of a truism, as we gain more and more 
experience working with people in a participative 

The establishment of meaningful goals and objectives 
is an important function of any participative concept. 
"Some goals are set by the individuals pursuing them, 
are set with the participation of others, some are set 
exclusively by others. Generally speaking, individuals 
most actively pursue the goals they set themselves. 
Invo.ivement in goal setting might take the form of units 
produced per day, per month, or per year. The 
establishment of a productivity objective, a reduced scrap 
level, or a reduced rework percentage can be goals also. 

What does all this involvement promote? If there is 
an acceptance and commitment by both participants and 
management, a climate of cooperation, understanding and 
genuine teamwork will prevail. To be successful in 
promoting teamwork in an organization, members of the group 
must perceive that higher level management is really 
interested in ar,d committed to fostering a climate based on 
open communication and mutual trust. 

When discussing a participative concept with 
individuals who are interested in what it is, or getting 
input from employees on what they think of a participate 
concept, an often-heard question is: "What's in it for the 
company? Obviously, there must be some benefits or you 
wouldn't be trying it!" 

The answer to this question is--Yes, there can be 
significant benefits, if recognized that a participate 
concept is an on-going process that changes as frequently 
as the weather--mainly because there are people involved. 
Because people are involved, and for the advantages to be 
fully realized, it must be approached from the standpoint 
that employees d^re truly ready and capable of making 
positive and constructive contribution. 


' '#.-V 



I- . 

■ 1. 



The benefits derived, in some instances, nay be 
measurable in dollars and cents, while, on the other side 
of the coin, benefits less measurable. but very 
significant. (lore specifically, measurable benefits 
reported by or^a.azations adopting a participative concept 
nave experienced the following: 

1 . Lower Absenteeism , which can be related to 
additional product shipped and on time 

2. Lower Turnover , less time and money spent 
on training new people. 

3. Higher Quality can be a result of the 
employees having a positive attitude about 
the importance of their task. A satisfied 
customer can generate repeat orders and job 

4. Meeting Product Targets in an atmosphere of 
enthusiasm and commitment. 

5. flchievement of Productivity Goals , which. 
In turn, lower product costs. The other 
side of the coin that is less measurable in 
teriTi of dollars and cents, but very 
significant from a humanistic standpoint, 
is that you capture a bigger part of the 
whole person. Perhaps the most important 
benefit of all this is simply the change in 
environment that, characteristically, 
occurs when a participate concept is 
attempted. The plant floor can become a 
strikingly different place from the days 
when work was just a routine grind. 

As previously mentioned, the development of the 
participate environment is not something that happens 
quickly—however internatiofK'^l competition and Japanese 
management styles have done much to bring about the 
awareness of the behavioral sciences effect on management 
styles in busines., today. As large numbers of 
organizations encounter successes in behavioral management 
applications, expertise will continue to be integrated in 
the organization. 






The industrial engine'^'' has traditior *lly over the 
years provided the interaction and communication link 
between management, technology and the worker. The role 
today has not changed; however, the process and culture 
have. As management gains greater familiarity with the 
behavioral sciences, management attitudes will change and 
thus enhance the cultural change. The behavioral scien<;es 
are now being regarded as results-getting tools and will 

thus be expected to produce hard, measurable improvement. 

The conflict between traditional industrial 
engineering and behavioral sciences and their practitioners 
is disappearing. In a quest for greater prciuctivity and 
efficiency, I.E. will put less emphasis on its traditional 
approach of work simplification and will view overall 
efficiency in light of what the behavioral sciences say. 
Worki-^ from the other direction, behavioral rcientists 
will need the skills of industrial engineering to develc i 
useful hard measurement. 

Change, and the management of change will be most 
significant in expediting the process. Because the 
development of high technology is so rapid, it is almost 
impossible to maintain the knowledge of the entployees to 
the level of technology. Because of this, two ,.ajor areas 
of tu$ organization need to be addressed: 

1. Organization has to be modified, changed, 
whatever to accept the new technology 

2. Retraining of the people relative to the 
new technology. 

The industrial engineering organization is best 
suited to deal with the change problem as it affects the 
organization. Recognizing that the industrial engineers 
must have prior training to become effective facilitators 
and thirxl party consultants--twr> approaches can be taken to 
address the organization needs for coping with 
technological change — ergo-the Socio-technical process. 

One approach might take in afsessing the needs of a 

new requirement, in comparison to the existing 

organization, is called functional organizational 

analysis. This approach utilizes a third party 

facilitator, but it brings into play all of tne 
participanLs or members of an existing organization. First 

take the existing organization and break it up by its 
specific func ions. 




As these particular function are broken out. they 
are further broken down by their specific responsibilities 
and the necessary tasks required to achieve these 
responsibilities. At this time, we are looking at an 
existing organization as opposed to the one that will be 
accepting the new technology. This gives the members o 
the organization an opportunity to see if, in fact, the 
responsibilities are still required for the job t!iat has be 
done today. 

As these functions are analyzed, we have the 
organization point out the skills necessary to accomplish 
the specific responsibilities that they have been 
analyzing. The members, at this point in time, build a 
skills inventory of the functional organization that 
presently exists. These are the skills which are deemed 
necessary to have the organization function properly in 
today's environment. This information also will provide a 
framework of reference when going through the 
organizational analysis needs for the new technology. 
After the functional organization has had an orportunity to 
be informed of the technology tliat is going to be 
introduced, it's possible to then go through a functional 
organizational analysis using the same type of approach 
that was used on the existing organization. The only 
difference here is working with the technologist. One can 
now define the new responsibilities of the functions that 
are going to be required to operate the new technology. 

Af these functions, responsibilities, and tasks are 
'defined, it's now possible to develop a needs analysis of 
the organization to satisfy the function with the new 
technology. At this point in time, the needs atialysis, or 
skills required can be compared to the existing skills in 
the organization. Further analysis will determine skills 
which are now obsolete, which have to be retrained, or new 
skills that are necessary to accomplish the operation of 
the function with the new technology introduced. 

The organization should also recognize, at this 
time, the new pkills necessary or new tasks that didn't 
exist before. The process allows the review of the 
organization and the required skill tasks that existed 
before to s^e where there are possibilities for job 
blending, ikiils blending, and areas where one can build on 
existing i'^lls. where there is a natural progression. If 
there is no natural progression, then it's ccing to be 
necessary to establish programs that can retrain the 
individuals so that they can, in ^act, function within the 
new organization and the technology that has bean 



With this type of information, it's possible to 
st^rt the necessary training plans for retraining or 
training the organization to accomplish the end result, 
that being, the introduction of the new technology. 

The second major area to be addressed is the 
training of the people for the^ new technology. This 
training can be placed into two different categories: 

1. Training that is necessary when the 
technology involwes a large number of 
people, such as the implementation of a 
large integrated factory business systems 

2. Where the technology involves a small group 
of people, the training may be for a piece 
of automated equipment or for the 
implementation of a robotics system. 

The development of the training pr cess required 
that certain objectives be met: 

1. Training of the personnel to be involved in 
the technology must be enmass, thus 
allowing a snort cycle time from the 
training to the application of the 

2. Training must be proficiency bised, not 
pass/fail, so as to ensure expected 
performance . 

3. Training must have "technologist" and 
"user" interaction and involvement. 

4. Training documentation must be modular and 

5. Training must not be a burden to the "user". 

6. Training and technology must have 
subsequent "user" ownership and control. 



Although the training involves different numbers of 
people, there is a conunonality between both types of 
training. This coirmonality is that both must have 
involvement of the organization and the users that will be 
trained. The training process is called the 
socio- technical integration process. It involves the 
introduction of the new technology to the factory floor and 
is a three-part process: 

1. Documentation of the technology. 

2. Development and training of the master 

3. Process is the actual training of the 
people themselves in the new technology. 

The documentation of the first step covers the 
documentation of the new technology so that it can be 
interpret'»d into user manuals or nianuals that will be used 
by the people in application of the new technology. These 
manuals are developed in modular form in that they can be 
broken down by functions and tasks which yill be performed. 

The user manuals become the text which will be used 

during the teaching phase, lesson plans are developed in 

the same modular man.ier as the manual. These lesson plans 
are developed to teach each module. 

The second area deals with the training of master 
trainers, who will subsequently train the trainers. These 
master trainers, for the most part, are those technologists 
who developed the technology. The technologists use the 
structured lesson plans when teaching the trainers, it 
structures them and provides assurance that the same 
information that they are teaching is taught the same way 
time and time again. 

Because these technologists are not professional 
teachers, it is necessary to put them through a training 
program themselves, so that they, in turn, can teach the 
trainers. This program is a two-step process, the first 
beirig a two-day program that deals with the principles and 
theory of how to teach in an industrial organization. 



;^l»N^-(.:i^V^ . "^ 


The second step of the teaching process is a 
training program which deals with the actual hands-on 
teaching of one's peers in a two-day training program. 
After the individual has done some teaching, they hav»: an 
opportunity to critique themselves, with an instructor, -nd 
discover their areas of skills or weaknesses. These becoi.Te 
the areas of concentration during the teaching process. 

Prior to the training of the trainers by the master 
trainers, the trainers also receive this same type of 
how-tc-teach training. This process was designed because 
of the large numbers of people to be trained. It was 
decided that the proper way to do this would be to train 
the supervisors of the subsequent users. These are the 
people that will actually be applying the new technology. 

By using the supervisors as the trainers, it allows 
you to get the pyramid effect relative to the teaching of a 
large number of people in a short period of time simply 
because each of the supervisors would then, in turn, teach 
their own people. Because the supervisors and their 
employees have to maintain the normal everyday work, the 
training has to be in the total control by supervisors. By 
giving control of the teaching process to the supervisor, 
it allows thetr. to teach when it best suits them. This 
allows the supervisors to establish the priorities of what 
modules they want taught first. Secondly, it also allows 
them, when they feel that someone hasn't properly learned a 
specific task, to take th^t person to the training area and 
give them some reinforcement or some retraining to maintain 
iob proficiency. 

This approach further provides the supervisor with 
the capabilities to subsequently retrain or train people 
for rotation and train new people coming into the 

fts previously mentioned, the process, both the 
functional organizational analysis and the subsequent 
training process, is developed around the user The 
purpose of that is to ensure the commitment of the user to 
the process. To have this commitme it, the ui^er must play 
an integral part in the development of the subseauent 
implementation of the technology. They must be tnere 
during the planning stages, not just handed a finished 
package at the end of the process. When using this 
approach, the implementation process can be accompli'^hed in 
a much shorter time, and is much smoother when going into 
the organization. 


-.,i» ._-;.. 


^•"■f?:!k\^. ~* ■ ~ ' ' {^\ 




The successful implementation and integration of new 
and advanced technology in the business organization today, 
is predicated on the cultural environment developed by the 
socio-technical component- balance. 

George L. Carter is Manager of Socio-technical Engineering 
at Lhe Westinghouse Manufacturing Systems and Tecnnolo ,y 
Center in Coluir'na, Maryland. He is responsible for the 
introduction and integration of advanced manufacturing 
technology (robotics, automatiufi aod computet ruanagement 
systems) into the workplace. 

Other positions he has held within Westinghouse include; 
but, are not limited to. Manager of Cor-^orate Industrial 
Engineering, Operations Manager and Plant Manager. 

A graduate of Baltimore University and Mount Saint Mary's 
College, he holds a B.S. in Industrial Management, Masters 
in Business Administration and Doctor of Laws degree. 



tJi?'^'*:'-^f^^- ' 

V ■ ' ' ' 14' 


Susman, Gerald I.. "Autonomy at W<irk", Praeger Publishiers, 



li^>*''?%v^, r 

■■ •'v >vS,«>. 



A Paper Submitted To The 

Track Entitletl, 

"Management Issues in High Technology" 

at the 

NASA Conference on R&D Productivity 

September lO-U, 1985 

Houston, Texas 

Submitted by: 

Rolf T. Wigand, Ph. D. 
School of Public Affairs 
Arizona State University 
Tempe, AZ 85287 


The author gratefully acknowledges the comments received from Dr. 
Dieter von Sanden, Member, Board of Directors, and Director, Corporate 
Communication Technology Research, Siemens AG, Munich, Federal Republic 
of Germany. 






Rolf T. Wigand, School of Public Affairs 
Arizon- State University, Tempe, Arizona 


The author reviews the work environment surrounding integrated 
office systems. He synthesizes the known effects of automated office 
technologies and reviews their known impact on work efficiency. These 
effects are explored with regard to their impact on networks, work 
flow/processes, as well as organizational structure and power. 
Particular emphasis is given to structural changes due to the 
introduction of newer information technologies in organizations. The 
new information technologies have restr'ctured the average 
organization's middle ranks and, as a consequence, they have shrunk 
drastically. Organizational pyramids have flattened with fewer levels 
since executives have realized that they can get ahold of the needed 
information via the new technologies o,uicker and directly and do not 
have to rely on middle-level managers, The author stresses the point 
that power shifts are typically accompanied with the introduction of 
these technologies resulting in the generation of a new form of 
organizational power. These effects and trends can be seen as an 
evolutionary step toward more flexible, decentralized and less defined 
organizational structures, resulting in a fluid organizational structure 
whose design is based on the changing needs of the organization. 


Office automation conjures images of legions of secretaries 
typing away in front of video display terminals. But the term implies 
no longer just word processing. Instead it is conceived today as a web 
of new technologies, including computers, PBXs, voice mail, electronic 
messaging, facsimile devices, conferencing systems, and others. The new 

*The author gratefully acknowledges the comments received from Dr. 
Dieter von Sanden, Meml'-er, Board of Directors, and Director, Corporate 
Communication Technology Research, Siemens AG, Munich, Federal Republic 
of Germany. 



muo >A. . '-V 

information technologies affect the access to and then flow and 
distribution of information throughout an organization and, eventually, 
the resulting changes can lead to crastic changes in the organizational 
structure itself. 

It is fairly recent that organizations have begun to view office 
automation efforts in a comprehensive context instead of in terms of 
individual technologies. Although approximately 87 percent of 89 
companies in nine countries stated that computerized information flows 
play a very important role in at least one corporate activity and even 
though the extent of computerized information use is expected to rise to 
92 percent by 193h (34, p. 67], several authors [e. g., 18, p. 57; 41] 
claim that organizations have Lardly tapped five percent of the 
possibilities utilizing integrated information technologies in office 
settings. At stake for organizations is an opportunity to speed \io 
communications, reduce paper flow and achieve productivity gains 
throughout the entire organization. Consequently, the integrated 
automated office is fast becoming a reality. It is quite obvious that 
only by integrating various office automation technologies into one 
coordinated system — an o rchestra t ed system — will today's office be truly 
efficient and effective. There are three broad classes of integration 
that can be identified: (a) the integration of information, (b) the 
integration of user interfaces, and (c) the integration across systems. 

A well-planned, integrated office automation system demonstrates 
five characteristics: 

1. It incorporates executives, managers, and support 

2. It transmits reliably fully integrated electronic 
documents — text, data, voice, as well as pictures — and 
stores them electronically across systems at high 
speeds, be it local or otherwise. 

3. It is a host-bated system designed around a central 
computer and a common architecture. As a consequence, 
end-users can gain easy access to common data bases, 
edit documents and transfer them between similar and 
dissimilar devices, including eouipment from different 
manufacturers . 

4. It is a real-time management information system, 
giving its u.sers access to ail information in the 
orgaiization and enables them co transmit information 
to their colleagues, wherever they may be. 

5. It has flexible conf i gura t ioi.s , an open architecture, 
transparent connectivity, compatibility, intelligence, 
f u 1 1 - tunc t iona I capability and, when appropriate, 
desktop functionality. 


• "•■'*2!5p 

Y <^''-^'^^-^ ■' 


Within just the last two years, a number of technologies have 
become available on the market that iii toto seem to make the integrated 
office come true. Ihey range from voice mail systems via local area 
networks to integrated communication satellite services. It seems that 
the growing sophistication of networks is currently the most important 
element that will make the integrated office a reality. Without such 
networks few of the real benefits of the integrated office can be 

Much of the needed technology exists today, although refinements 
are necessary to tie the individual elements of the automated system 
together. An integrated office system is more than just the sum of its 
parts and it is specifically the (Applications that matter most, not the 
devices. Planners and users alike must learn to understand that they 
are not a single-cell unit, but they are part of an interdependent 
community. In this sense then an integrated office syscem is not 
defined by its hardware, but by its software. This software in turn 
should allow dissimilar devices to communicate among each other and make 
the system easy to use. In accordance with this line of thinking, there 
appears to be a growing recognition that the management of information 
and effective communication are aa i.nportant, if not more so, to 
organizational success as are p'^oducts and services. 

The Work Environment Surrounding Integratea Office Systems 

One can identify three classifications of office workers as 
prospective users of an integrated office system: (1) those making 
limited transactions account for 33 percent of the workforce, (2) full- 
function professionals amount to 27 percent, and (3) specialized 
professionals constitute five percent of the workforce. The remaining 
35 percent are white-collar workers who are not potential system users 
due to the specialized nature of their organizational activities [41, p. 
28]. In 1983 there were 53 million white-collar workers in the United 
States and 35 million were potential users of integrated offices 
systems, but less than five million had only desktop devices available 
CO them (41, p. 28]. Even today *Tuly infregrated desktop systems have 
enjoyed very little penetration in most organizations, a shortcoming 
chat is likely to change rather drastically with the massive advent of 
personal computers on workers' desktops. One must consider, however, 
that while striving toward? automated integration, such desktop devices 
must be linked at least via a local area network and ideally linked with 
a central mainframe computer in larger organizational settings. Within 
this environment office systems are moving toward the integration of any 
function that can be carried out by a number of technologies. 

The market for telecommunications equipment and services in the 
United States is certainly a most promising one. It is expected to 
reach $150 billion cnnually by 1987, a 50 percent increase over current 
spending [19, p. 22]. Sales of communications technology hardware alone 
has reached $60 billion in 1983 according to Arthur D. Little estimates 








and will increase to $90 billion annually by 1988 [44, p. 59]. 
Information processing has also grown into an enormous industry, 
accounting for $33 billion in services in 1983 — the last year for which 
figures are available — and it is projected to account for $88 billion in 
1988 (17, p. 57). Office automation — according to some estimates — has 

,, reached only a small segment of the 14 million U. S. businesses with 

revenues less th^n $10 million (23, p. 126]. Office costs have been 

\ estimated to ircrease at a compounded annual growth rate of 15 percent 

(1. p. 261. 

Such market potential is also reflected in the various activities 

which typically can benefit from the use of communications equipment. 

i"» For example, Xerox reports that in 1981 a total ot 850 billion pages of 

business documents were produced in the United States compared to an 

anticipated 1.4 trillion pages in 1985 (36]. This latter figure is 

expected to be produced and handled by centralized in-house and contract 

printing (38 X) , by centralized data processing cencers (34 X) , by 

office machines (15 X) and by distributed forms of data processing 

(132). Over twenty-one trillion pages of paper are now stored in the 

United States based on studies by Frost & Sullivan (28, p. 53]. The 

r same report claims that for earh of. the eighteenmillion U. S. tffice 

"i| workers, there are four file drawers with 4,500 documents a piece. 

'jv- Furthermore, office workers are creating new documents at the rate of 

^'^ one million per minute. This translates to the annral creation of 4,000 

.f new documents per office worker. At the same time this individual files 

ten pieces of paper each day, amounting to a daily total of 180 million 

pages. Within the time span of a year, this makes up 46 billion, 800 

million pages. There is no doubt that paper is still the number one 

information-carrying medium in the world despite recent dramatic inroads 

of office automation systems. 

When analyzing how managers tend to spend their available time 
gives additional understaiding about the potential of integrated office 
automation systems. Managers in over 30 studies — too numerous to detail 
here — spent when averaged 38.29 percent (SD " 17.45) of their time in 
face-to-face settings, 10.50 percent (SD " 8.00) of their time is spent 
on the phone, 14.58 percent (SD = 7.73) reading and 14.57 percent (SD = 
5.95) of their time is occupied by writing. Similar studies report that 
46 percent of the manager's time is spent in meetings or with telephone 
calls, 25 percent is spent with administrative tasks, 13 percent with 
document creation activities and the remaining 16 percent fall into the 
analysis and other category. Others have reported that only seven 
percent of the manap"i's time is spent with primary functions (decision- 
making) and that 78 percent of available time is spent with various 
communicacion activities such as receiving, storing, filing, retrieving 
or transmitting information [Cf., e. g., 10, p. 40]. Professionals spend 
about 20 to 30 percent of their day just searching for information (43, 
p. 78]. Secretaries spend 25 percent of their time away from their desks 
and about 18 percent waiting for work, i. e. they are either unavailable 
to those they 'ipport or just idle (43, p. 78]. Others claim that only 
20 percent of a secretary's day is spent typing. Secretaries are 




■- '#i~i.- 

^r^^v--." ■ ■ ■ ■ (*; 

interrupted approximately 45 times a day (43, p. 79]. Such data should 
be seen in light of the fact that the cost of dictating and transcribing 
the "average" letter amounts now to $8.10, a 6.6 percent increase over 
the 1983 cost of $7.60, according to the annual survey of the Dartnell 
Institute of Business Research [8]. It is still quite conunon Chat the 
majority of office tasks are not analyted and proceduralized and, many 
times, when procedures exist they are eithc ignored or out of date. 
Quite frequently, people in the office perform tasks which are clearly 
marked for lower paid personnel as can be observed when a manager or 
professional stands in line waiting for the photocopying machine. 

From the results of a recent study on the contents of telephone 
messages conducted by AT&T [33, p. 36] we know that less than ten 
percent of the messages are "complete." except for the category of "no 
messi je" calls, these phone calls will trigger at least one more call, 
even though it is highly probable that a complete message would have 
enjoyed higher productivity results. This AT&T study found that in 
terms of mebdage content 46 percent left name and number, 26 perceni; 
left name, number and purpose, ten percent left their name only, nine 
percent left a complete message and nine percent left no message. It 
should be noted, however, that 82 percent of these telephone messages 
constitute really no more than a "call report" due to the interpersonal 
barriers of third-party message taking. Similar results have been 
reported by Hirschberg, et al. [9]. Asten [1, p. 31] estimates that U. 
S. busiiess has currently a $15 billion "pink slip problem" (for 
telephone messages) and that 64 percent of the many millions of return 
telephone calls made are actually displaceable. 

Technology has been advancing manufacturing productivity for 
centuries, but relatively little has been accomplished for the 
information and services sectors of U. S. industry. These sectors 
constitute over 50 percent of all four labor sectors an^ the 
'information business' accounts for 60 percent of the nation's economy 
[Cf., e. g., 22, p. 30]. With the information and services sectors 
being the largest onas, it is here where labor saving nethods will pay 
off the most. Numerous productivity-enhancing office automation 
technologies have already or will produce large advances in 
productivity. One estimate that information flows are increasing at an 
annual race of ten percent and since over half of the U. S. work force 
is? now processing information, it makes eminently sense to automate 
increasingly information processing and transmission functions. 
Otherwise, we are likely to choke on information overload as the 
statis^ic8 below may suggest. 

It is most surprising that investments in office workers have not 
been considerably higher. Blue-collar productivity increased more than 
80 percent during the 1960s, yet U. S. white-collar productivity 
increased merely fout percent during the same time span (40, p. 64n). 
It is generally estimated that $30,000 in start-up equipment, services, 
and training are required to support just one knowledge worker [13, p. 
69], the term frequently used to denote managers, administrators and 





professionals as a group. Poppe 1 (29, p. 147] reports from a Booz, 
Allen & Hamilton study that more than $1 trillion went for salarieH and 
support of white-collar workers in 1982 and $600 billion of this figure 
went to compensate knowledge workers. Some experts predict that by the 
year 2000, 119 million people — amounting to nearly 90 percent of the U. 
S. work force — will be white-collar workers. By 1990 at least hr.f of 
all office workers in the U. S. will use a word processor, a data 
processor, or some varirnt thereof [24, p. 72]. Secretaries and typists 
represent only 8.7 percent of tctal office salary costs, while 
professionals and managers make up 68 percent fl4]. It is estimaced 
that word processing systems could eliminate 3.8 percent of all office 
costs if they replaced every typewriter, the savings yielded from 
automating professionals and managers wculd be substantially higher. 
According to Kearns [14], interpersonal communication, analysis and 
decision-making make up 47.2 percent of all office expenditures. In 
spite of those figures though, 'oy early 1980, only J 5 percent of major 
U. S. businesses reported having more than five qualifii>d, full-time 
office automation professionals on their staffs l29, p. 154], As the 
cost of automated office equipment has declined and becoce more accepted 
in the executive suite, it appears that the race is on to uiake white- 
collar workers considerably more productive. 

Work Efficiency Due To Office Automation Efforts 

A wide variety of office automation technologies are available to 
be used for various organizr«:ional activities. They could be classified 
into four phases of product preparation, i. e. input (conversion of 
ideas or thoughts into verbal or written communications, production 
(processing or manipulating ideas or thoughts created ^nd/or stored 
during the input phase), output (the generation of an electronic, 
optical, or hard copy document) and distribution (transmission or 
movement of output documents or data). The tools in this sweeping 
electronic upheaval are the interlocking parts of a communication 
network '^hat was inconceivable just a decade ago. These technologies 
include the following: 

Communication satellites 

Computer conferencing technology 

Computers (includini^ personal computers, desktop workstations and 

terminals ) 
Electronic mail 

Optical character recognition 

PBXs (telephones, including cellular and mobile telephones) 
Record? iranagement technology 
Teleconferencing technology 
Video disks 
Voice messaging, voice recognition and voice activated tecnp.ology 




-^ . • ■ ' " \4-' 

The limited space nere would not allow to address all known 
positive and negative effects for each of these technologies. In order 
to overcome this difficulty an effort is being uade to suaunarize these 
findings for office automation settings, especially integrated office 
automation, as much as possible. Rice and Bair [32, p. 189] present a 
broad summary of key productivity findings and benefits of office 
automation and grouped them into five areas: 

1. Control ; requiring less information to perform a 
task, better planning, providing a more effective 

2. Timing : reduced waiting time for a meeting to commence 
or for another department to respond to an inquiry; 
reduced time spent in decision-making, initiating 
action, or responding to the environment; increase 
flexibility of work schedule. Uhlig et al. [38, 
chapter 3] elaborate on the role of these components 
of a cybernetic communication system in improving 
productivity. The other three areas involve the form, 
transformation, and by-products of communication 

3. Automatio n: the replacement or elimination of manual 
processes, such as constant revision of mailing lists, 
that do not contribute to increased effectiveness. ... 

4. Media transformations : time, energy, and errors in 
transferring information from one medium to another 
may be reduced. For example, a company memo may pass 
through many media--oral, tape, typewritten, 
handwritten, revision, typewritten final, photocopies, 
and mail — b>?fore the content is entered into someone's 

5. Shadow function^ : unforseen, unpredictable, time- 
consuming activities that are associated with 
accomplishing any task, but do not contribute to 
productivity, including telephone tag and unsuccessful 
attempts to retrieve information from a personal 
tile, ... 

These class categories are, undoubtedly, useful. On the other 
hand, they are at such a high level of abstraction that they are of 
little direct utility for the practitioner. An attempt is made to 
summarize key work efficiency findings for office automation settings. 
They are presented in Table 1. These and other studies have investigated 
the impact of information technology upon organizational settings. 
Space limitations do not permit additional elaborations on those 
specific results. 



Work Efficiency Impacts of Office Automation Settings 

Research Reference 

Bair (1974) 

Bair (1980) 

Bamford (1978, 1979) 

BuUen et al. (1982) 
Canning (1978) 
Conrath & Bair (1974) 

Cramer & Fast (1982) 
Crawford (1982) 
Curley & Pyburn (1982) 

Dahl (1981) 

Dunlop ec al. '1980) 

Edwaris a978) 

EIU Infcrmatics (1982) 
Rank Xerox 

German Computer 

British Dept. of Ed- 
ucation & Science 

Engl and- various 

British travel 

Engel et al. (1979) 

Gardner (1981) 

Gutek (1982) 

Helmreich & Wimmer 

Johansen et al. (197<^) 

Kalthoff & Lee (1981) 

Work Efficiency Impact 

Changes in communication patterns: decreased 
face-to-face contact andincreased vertical 

Reduction of telephone shadow costs 
Decreased document turnaround time and in- 
creased document output 
Improved work flow with time saved 
Increased volume of communications 
Reduction in telephone and face-to-face com- 
munication; increased upward communication flow 
Considerable secr'^tarial time saved 
Decline in menial, clerical tasks 
Greater organizational productivity when seen 
as more than a way to increase secretarial out- 

Annual cost savings and 8% increase in knowl- 
edge worker time 

Striking productivity gains from 50-100+ per- 
cent in document preparation 

Change in working based on ability to work at 

Large reductions in telephcL.e use, 1 hour/day 

labor saved for attorneys 

Joint information sharing and editing 

50-100 X productivity gains in word processing 

after 6 months 

Many sites did not reach co£t-justifying word 

processing productivity levels 

Word processing output rose 150 X 

Word processing output rose 150 X 
5-25 X knowledge worker and 15-35 % secretarial 
time saved 

Decreased turnaround time in document prepara- 
tion and scheduling 

More creative, challenging, complex, organized 
High User Acceptance 

It takes longer to transmit the same amount of 
information in electronic print form than 
transmitting it verbally 

Use of micrographics system resulted in mini- 
mum of 30 X saving of annual operating costs ; 
reduction of 90 X or more in space require- 
ments; more rapid information retrieval (25 % 
to several hundred percent faster); increased 
file security and document integrity; substan- 




'..^V;>4>^-^ t. 


Table 1 — Continued 
Research Referfcace 

Work Zltic cy Impact 

Kaplan (1980) 

Kerr & Hiltz (19(j2) 

Leduc (1979; 

Lippitt et al. (1980) 
Miller & Nichols 

Melton (1981) 

Mertes (1981) 

National Archives and 
Records (1981) 
case processing 

electronic messaging 
optical character 
National Bureau of 
Standards (1980) 

Panko & Panko (1981) 

Picot & Reichwald 
(1984), Picot et 
al. (1982) 

tially decreased duplication and distribution 

Use of dictation equipment resulted in 6.23 to 
12 X time savings per day 

Agreement on the group's decision tends to be 
lower in computer conferencing groups. When 
groups have longer periods, the level of agree- 
ment seems to improve in such groups . 
Changes in coimnunication patterns: ability to 
work at home and increased vertical communica- 
ReducLion in memos and phone calls 

Improved jmmunication and decreased informa- 
tion float 

Cost and time savings via central library 
Concept, 3cope, and potential of office auto- 
mat' on ill-defined in government 
Reduced backlogs, more timely and accurate 

Too few terminals — no benefits identified 
More cost-beneficial than word processing 

Use of stand-alone display text editors result- 
ed in an average of 75 X to 133 X productivity 
improvement if used for original and revision 
typing; average 104 % to 181 X if used for 
revision only 

Use of word processing impact printer (letter 
quality) was 148 X to 533 X faster than elec- 
tric typewriters' capacity 
Use of data processing printer (matrix) was 
1,184 X to 3,554 X faster than electric type- 
writer for a matrix, 40-120 cps printer; 4,813% 
to 6,417 X faster than electric typewriter for 
a 150 to 2,000 lines per minute (line) data 
processing printer. 

Use of micrographics enjoyed a 25 % savings in 
retrieval/access time 

Use of facsimile (fax) — the transmitision of an 
8 1/2" by 11" page took 30 sec. to 6 tcin. ; when 
6 min. compared to 2 days for mail to arrive, 
fax was 480 times faster 

Many perceived benefits — more for managers and 
professionals using system directly than for 
secretaries; highest benefit for long-distance 

A tension exists between a favorable attitude 
toward technological innovations in offices and 
skeptical view of the specific personal conse- 
quences to be faced when technological change 
in offices occurs. On the one hand, a large 




«ki V '. », 


Table 1 — Continued 
Research Reference 

Work Efficiency Impact 

Rice & Case (1982) 

Rice et al. (1984, p. 

Rice et al. (1984, 
pp. 137-140) 

Steinfield (1983) 

Steward (1983) 

Talbot et al. (1982) 
Tapscott (1982) 

Tucker (1982) 

Uhlig et al. (1979) 
Witte (1980, 1977) 

Witte (1976), 

Szyperski (1979) 
Yankee Group (1978) 

Zouks (Cited in "Com- 
puters Rescue," 1980) 

majority (80 X) articulates a positive opinion 
on new office technology; on the other hand, 
wide-spread fear (60 Z( of unfavorable conse- 
quences for own work situation exists. 
Substitutability and choice of communication 
channels were studied 

Reduction in paper and telephone traffic; in- 
creased work quantity and quality; but depends 
on "'<edia style" 

For electronic messaging, benefits: permanent, 
searchable record; fewer meetings are needed; 
control of time to respond; independent of time 
zones and geography; quick delivery; easy to 
distribute widely; speedy response; upward com- 
munication is encouraged; can serve as channel 
substitute; potential for reduction in travel; 
medium for creation, transmission and receipt 
of message are the same; fewer nonverbal con- 
straints; - drawbacks: potential for informa- 
tion overload, misunderstandings can occur, 
people may be inhibited by machine, much use- 
less information can be distributed due to 
ease of use 

As compared to computer conferencing, it takes 
less time to arrive at a decision in face-to- 
face groups. When the studies had time limits, 
this fact also takes the form of less consensus 
because the computer conferencing groups have 
had less effective time in which to consider 
the problem. 

More timely and accurate information, increased 

13 X daily time saving; 35 X increased document 

Less time in communication activities 
Changes in communication patterns: decreased 
telephone usage and meetings 

Implementation failure primarily due to short- 
ened initial analyses phase 
Reduction of telephone shadow costs 
By means of new communication technologies, de- 
centralized aatonomous groups could pursue 
their work effectively without risking organi- 
zational disintegration 

New structural configurations for innovative 
organizational decision-making could emerge 
Use of electronic mail resulted in a 40 % re- 
duction in dissemination time 

Use of personal (professional) terminal result- 
ed in 50 X productivity rise 

Note: Some of EHe above references and impacts were first compiled by 
Rice and Bair (198^, pp. 211-213). Due to space limitacions only the 
newly added research references are cited in the reference section. 




Goldfield (7, p. 66] classifies work efficiency and office 
productivity studies according to the natur3 of the issues involved and 
the associated problems by organizational layer, as represented in Table 


Classification of Work Efficiency and Office Productivity Studies 


Layer Description 


Top Executives Most expensive, greatest 
Conceptual thinkers 
Decisions critical for strategy highest taboo area 


High salaries 

Control of Operations 
Decisions critical for 

Difficult to identify 

Tasks varied 
Stiff resistance to 


Professionals New target group 
Technical skills 

Qualification critical, 

Task specialized 

Support Staff Easiest to quantify and 
cost justify 
Most frequently automated 

Often least significant 
savings opportunity 

Of particular concern to those interested in integrated office 
automation systems are various structural changes and changes that can 
occur with organizational power that are due to the introduction of 
newer information technologies.. 

Squeezing the 'Organizational Sponge': Structural Changes Due To 
Information Technology 

The computer has changed the offrce environment just like the 
automobile has changed the city. Integrated office automation is a 
revolutionary development in the transfer and general dissemination of 
information and data. It clearly has the power to shake an organization 
from th<i top to its very foundation. An effective integrated office 
automation system may result in the restructuring of one or several 


■■- ''iCfsW'V* 


organizaL. jnal units ranging from minor adjustments to the eradi ':ation 
of a unit. 

Several authors have argued that new structural configurations 
could emerge within organizations employing newer information 
technologies [37, 47, 50]. Already as far back as 1958, Leavitt and 
Whisler predicted that the design of large-scale hierarchies will be 
changed from a pyramid toward a bell formation in line with changing 
distributions of power and authority within the organization. Most 
organizations enjoyed a relatively rapid growth since World War II and 
with such growth, middle management grew even faster. In the 19708, the 
,, U. S. economy absorbed a rapidly expanding labor force and created 19 

million new jobs, whereas most western European nations' economies were 
•■, at a standstill or suffered declines in the post-OPEC era. In effect, 

' the nation had pur the Baby Boom generation, born in the 1940s and 

19508, to work and these Baby Boomers are now reaching middle age. 
Working women at an age of 33 and older have been instrumental in 
allowing business to expand since World War II. Women have seized two- 
thirds of the more than 20 million jobs created in the past decade, 
accelerating the shift from manufacturing to services [49, p. 80]. 
i Midd le management together with professionals and technical people 

i]^^ constituted the fastest growing sector in the occupational groups in the 

, last 30 years. 

;. Typically middle management was to take policy decisions received 

t from top management and translate them into profitable revenues. Staff- 

'.^ level middle management was expected to advise their superiors on 

■; marketing, strategic planning, manufacturing and engineering issues. At 

an increasing pace middle managers became the collectors of information 

, that they in turn analyzed, interpreted and then moved along to top 

/^ executives. From these type of activities the middle manager came to 

dominate line operations. During the last few years, however, it appears 

that the function of middle management has changed and that mnch of this 

change is due to recently introduced information technology [3, 5, 21, 

23, 42]. This technology is rapidly reorganizing the kind of work 

people do and, at the same time, reorganizing our organizations as well. 

New patterns of communication and interaction become possible and, as a 

consequence, our established communication networks within organizations 

change [cf., e. g., 26]. The newer information technologies make it 

oossible that a greater amount of information is made available to more 

and more people in an increasingly timely manner. Such changes, no 

doubt, can influence decision-making patterns and can result in 

organizational power plays. 

Rather recently, at least since 1982, the middle manager's role 
was turned upside down. Middle management has been reduced in 
significance and numbers because its basic function of communication 
has — at least in part — been assumed by machines. The vast in-roads made 
by the newer information technologies appear to have restructured the 
average organization's middle ranks and, as a consequence, they have 
shrunk drastically. Maybe the first results of chis radical phenomenon 




■■ - '«*«"■&:! '. 

could be ob8«rved after the most recent recession in that many laid-off 
middle managers — until recently a rather stable and secure position 
within the organizational hierarchy, earning anywhere between $25,000 to 
$80,000 — were not rehired. 

Frequently, top executives viewed middle management as an 
organizational sponge sitting 8olidl:y in the middle of the 
organizational pyramid. If one pushed the elastic sponge a little bit 
from above or below, not much happened. But if one tried harder and 
squeezed the sponge some action occurred, although — just as with real- 
life uponges — once the pressure subsided, sponges have a tendency to 
assume their old shape, size and volume. Now — it appears — new means 
have arrived to deal with this phenomenon. 

As more and more top executives recognize that much of the 
information previously collected by middle managers can be secured 
quicker, cheaper and more thoroughly by computer-based devices, they 
have started to view much of the middle management layer as redundant. 
Many now view these information gathering, analyzing and interpreting 
tasks traditionally carried out by middle managers as considerable cost 
centers, high in overhead and making relatively small, directly- 
attributable contributions to profits. They are realizing in accordance 
with their Japanese competitors that less means more. Many bureaucratic 
functions can be replaced by information technology. In addition, since 
much interaction between a manager and his/her subordinates can be 
carried out nonsimul taneously via electronic media, a manager can take 
on direct responsibility over several more individuals. This, of 
course, could result in fewer managers at the intermediate level, but it 
might imply additional positions at the remaining levels. Jennings 
[21] claims that one-third of the 100 largest industrial companies are 
reducing management and that there are clear signs that other 
corporations will follow. Analogously, during the first quarter of 1983 
the U. S. Bureau of Labor Statistics reported that unemployment among 
managers and administrators in non-farm industries was the highest since 
World War II. The latter report did not even include those managers who 
took advantage of early retirement programs and similar incentives. 

Even old- line companies have made cuts in their middle-management 
staffs: Firestone and Crown Zellerbach, e. g., cut their middle- 
management staff 20 percent, Chrysler did the same with 40 percent [21, 
p. 52]. Similarly, a few leading edge companies have reported 
significant structural changes due to new information technology: The 
chairman of Hercules, Inc. cut the levels of management between himself 
and Hercules plant foremen from a dozen to six or seven. FMC 
corporation reports its amalgamation of sales districts and, in many 
cases, cut out a level of management. FMC installed a voice-mail system 
and the resulting reduction in cost for long-distance calling alone paid 
for the voice-mail system. Citicorp introduced a sophisticated 
information system to improve customer service and make account and 
market information available to its corporate clients more quickly. 
This step allowed Citicorp to reduce its staff — 70 percent of wnom were 




■« Vi » 


formerly clerical — from 2,650 three years ago to 2,150. This bank has 
now been able to move its customer service operation closer to its 
clients since information can now be transmitted electronically. More 
than 90 percent of its employees were based in New York in 1979 vs. only 
> 50 percent today [23, p. 118]. Similarly, General Electric laid off 

eight percent of its white-collar work force in November 1984 at its 
vast Appliance Park Complex in Louisville amd GTE Corporation, whose 
total payroll has fallen to 183,000 from 200,000 in the past two-and- 
onehalf years, has been cutting layers of management and combining jobs 
to boost white-collar productivity (45). 

I It has been estimated that by the year 2000, computerization will 

have caused only a moderate reduction in total jobs. The likely losers, 

^ ' however, are expected tu be white-collar clerical and managerial 

positions. Clerical jobs are said to decline from 17.8 percent of the 
work force (in 1978) to 11.5 percent in 2000. Although there will be 
major growth in the professional sector, the need for managers vill fall 
off [16]. 

'f Together with economic necessity and technological forces it 

; |. appears that middle management is affected by such developments in a 

> ^ number of ways. The corporate structure tends to be changing in that 

*-f. broader information gathering can be accommodated and that data can flow 

to top executives and managers directly without the editing, monitoring, 

interpreting carried out by middle managers. Many times, the analysis 

of such activities provide an opportunity to examine what workers 

actually do. A redesigned organization is likely to pull together and 

merge under just one organizational umbrella department such units as 

; office automation, management information systems (MIS), data 

processing, word processing and telecommunications. These efforts may 

result in the creation of a new group of information managers. They, in 

turn, would be responsible for the use and pro^tfction of valuable data 

and, in a sense, they would become the custodian of organizational 

information. Information centers, as such newly created organizations 

are called, are growing at a rapid rate. Of 160 large North American 

corporations questioned in mid-1984 by the Diebold Group Inc., 80 

percent stated that they had set up information centers this year (up 

from 67 percent in 1983) [23, p. 124]. Numerous industry experts predict 

■^ that end-user computing is going to be the dominating means of providing 

information support and will result gradually in a major organizational 


As a result of these changes, the management pyramid has begun to 
flatten with fewer levels and, as a consequence, emp/loyees make more 
lateral moves and their expectations tend to be lowered since fewer 
organizational layers means fewer chances for promotion. Any changes in 
the organizational structure are today largely perceived as a by-product 
of the new information technologies. Eventually the structure of 
departments and divisions will change as they increasingly share 
information for operations and decision-making. One data base, e. g,, 
. " may support several departments and others could be merged and 







integrated to make full use of relational data base features. These 
tretids may suggest that middle management positions become less secure 
and more competitive. Furthermore, organizational structures will 
experience a gradual shift from the typical pyramid to a shape 
resembling a diamond. The clerical function is likely to be distributed 
organizat ion-wide . 

A New Organizational Power 

The effect of automation on organizational power can be observed 
ir a number of ways [50]. As more and more timely information becomes 
available to a larger number of people, this increased availability of 
information can result in changes in the decision-making patterns and 
power bases of organizations. Organizational power can be defined as 
the ability to exert influence and bring about desired outcomes, 
including {!) the ability to resolve uncertainty and solve problems, (2) 
the power of expertise and (3) diminishing job specialization fCf., e. 
g., 2. p. 34; 35, pp. 306-328; 25, pp. 2-3]. With the newer 
information technologies available in organizations, political power 
associated with the control of information becomes more and more 
distributed in most organizations. Some, however, experience the 
opposite for certain technologies, i. e. control can become more 
centralized. Access to information via these new technologies makes 
greater centralization of power and control in organizations possible, 
since top management is enabled to make decisions quicker and by 
themselves that would otherwise have to be delegated. 

Routine activities can also be relatea to decision-making power 
[2]. Task routinization reduces uncertainty in a job and. in turn, 
provides an avenue to reduce uncertainty for someone carrying out this 
task. Furthermore, task routinization makes posiible the reduction of 
unknown entities, reduces the complexity of the task at hand, requires 
fewer skills and makes these executing the task more replaceable. It 
follows then that those who carry out routinized tasks have less power 
with limited or no decision-making authority in their work environment. 
These are usually also the reasons why employees ma" resist office 
automation efforts. In part depending on the structure of the task and 
the degree of required specialization, a secretary may, e. g., resist a 
changs to word processing when this individual was accustomed to 
performing variable tasks. 

Frequently, the strongest resisters when a new technology or 
procedure is introduced are almost always the middle managers. There 
are good reasons for this though: they equate their power with the 
number of people they supervise and anything that changes this is likely 
to threaten them. They are often left out of direct involvement with 
the new technology and, as a consequence, they do not experience the 
direct benefits themselves. Usually they have been promoted Decause 
they understand the current procedures better than their subordinates. 
Kearns [13, p. 70] claims that "The trouble with a lot ' f white-collar 


tofcs^,*kisr-%^'':.. /-., 

people is that they think the reason they are employed is that they are 
experts ." 

On the other hand, the newer information technology can 
strengthen the power of individuals or entire orf.anizational units. In 
part depending on how the equipment is utilized, individuals or units 
could be perceived as being irreplaceable which always was viewed as an 
iipportant source of power. Some individuals or units may be able to use 
information technologies while others are not or have not yet received 
such technology. Until the technology becomes widely available as a 
resource, the parties involved can enjoy a certain degree of leverage or 
power due to the access to and control of the technology. 

Such information technology makes it possible for individuals or 
units to carry out many more tasks. As ii result, the degree of 
specialization is reduced in many organizations and individuals or units 
can take on additional responsibilities which may imply a loss or 
possibly an increase of organization power. Furthermore, such 
developments are likely to make organizational units more interdependent 
with others which can be accommodated more easily and can make increased 
information exchange and sharing possible due to the available 
information technology. Overall, as specialization is reduced, this 
should have positive results on various organizational conflict 
situations and power struggles since information is now largely 
distributed as opposed to being available to only certain individuals or 
unite. It will be more and more difficult to control absolute, i. e. 
undistributed, information when on-line information systems and joint 
databases are used. 

Conclusio n 

As organization charts are redesigned, as chains of command are 
simplified and as the new structures are implemented, the balance of 
power is changing and shifting in our organizations largely as a result 
to information technology. Rigid, hierarchical structures are redesigned 
resulting in leaner, more flexible and responsive organizations with 
fewer management levels and more direct infoimation exchange between the 
top and bottom layer of the organization. Organizations are ci .gating 
environments in which performance determines compensation and a relative 
high degree of freedom to improve performance becomes the psychological 

The phenomenon described here '.nay be observable only in a few 
large organizations so far, but it is expected to hit many more within 
the next few years as they automate tlieir offices. This trend can be 
seen as an evolutionary step toward more flexible, decentralized and 
less defined structures, resulting in a fluid organizational structure 
whose design is based on the changing needs of the organization. IBM, 
e. g., is one organization that has long had this type of organization 
in place. Organizational structures serve both to differentiate and the 
integrate the components of complex organizations. Organizations do not 




, *,.'/> 

achieve and maintain effectiveness by technology alone. Studies have 
shown that the effectiveness of an organization is contingent upon the 
degree of fit it achieves between the technology and its structural 
design. Since organizations are not things, but people, it is 
ultimitely people who will be changed by the impact of information 
technologies. As organizational structures change, the distribution of 
power will change as well. This problem is unque^ t iona 1 1 y 
organizational and cultural within its respective setting. Deal and 
Kennedy [6] predict the arrival of the "atomized organization," meaning 
that telecommunication networks and common culture, i. e. not the 
organization chart, will link the "atoms." It seems that ways must oe 
found that improve the flow of information to the right people and to 
produce true entrepreneurial opportunities for managers who want such 
opportunities within their organizational framework. This implies that 
integiated information technology must be applied creatively to change 
the way organizations carry out their activities and use this technology 
to their own competitive advantage. In most organizations the technics' 
experts responsible for managing information systems have tended to 
treat office administration less as a strategic weapon to win a 
competitive advantage and more as a productivity tool to improve the 
performance of clerical and managerial workers. They have viewed office 
automation products as timply a means to computerize old procedures. 
Executives must analyze how they can use new information technologies to 
nelp restructure the organization and and pare down unnecessary 
organizational layers. The question to ask is, 'How would I change my 
organization to make it more effective with information technology?' 
rather than, 'How can I use office automation procedures to make my 
employees more productive?' In the past, many managers merely focused 
on making individuals more productive by overlaying electronic systems 
(e. g., word processors, ^ onal computers, EDP) onto the existing 
bureaucracy. Before we Ca.. address the iTsue of integrated office 
systems, however, we nust come to realize that an integrated system 
requires integrated management. An organization must first evaluate its 
structure, for the lack of integrated management will continue to affect 
the success of an integrated systeni after its installation. 
Organizations should not expect long-term gains when they understand the 
technology. Out they do not understand themselves. Integrated 
information technology systems applied in this suggested way are likely 
to turn out to be the single most important opportunity for private and 
public sector organizations until 1990. 



- '*-.'rt- 

Il^-*'*,^- ?^ .^-s. 




[ 1] Asten, K. J. Integrating conmunications in the automated office. 
The Office , June, 1984, pp. 26, 31, 66, 68, 73. 

[ 2] Collier, S. D. The effect of office automation on organizatio'\al 
power. The Office , 1984, 99(2), 34, 36, 44. 

[ 3] Communirations in dealing buoiness a potent new hand. Business 
Week, October 24, 1983, pp. 131-132. 

[ 4j Co-nputers rescue for travel agents. Business Week , April 7, 1980, 
pp. 81-82, 

[ 5) Computer shock hits the office. Business Week , August 8, 1983, 
pp. 46-49, 52-53. 

[ 6j Deal, T. E., & Kennedy, A. A. Corporate Cultures . Reading, MA: 
Addison-Wesley, 1982. 

[ 7] Goldfitid, R. J. Aiming OA towards the top. Modern Office 
Technology , 1985, 30(2), 55-57, 60, 62, 64, 66, 68. 

[ 8) High cost of communicating is getting higher. Modern Office 
Technology , October, 1984, p. 22. 

( 9] Hirschberg, L.; Bossaller, B. ; Kerns, P.; Dugan, D.; Sargent, L.; 
Feltman, .".; Wiseman, S. & Datton, W. The Notice Market: A 
Comparative Analysis of Electronic Message Systems . Los Angeles: 
University o£ Southern California, Annenberg School of Conmurica- 
tions, 1982. 

[10] Horwitt, E. Confronting the communications quandry. Bu siness 
Computer Systems , 1984, 3(9), 38-41, 44, 45. 

[11] Johansen, R. Teleconferencing and beyond: Communications in the 
office of the future . New York: McGraw-Hill, 1984. 

[12] Kaplan, H. M. Words , June-July, 1980, pp. 40-43. Cited in 
National Bureau of Standards, U. S. Department of Commerce, 
Computer science and technology: Guid ance on requirements 
analysis for office automation syste ms. (NBS Special Publication 
500-72. (C # 80-600179X Washington, DC: Superintendant of 
Documents, U. S. Government Printing Office, 1980, p. E5. 

[13] Kearns, D. T. Computers won't make a bad business good. Manage- 
ment Technology , 1984a, U12), 69-70. 

[14] Kearns, I. S. Cited in: "Office automation restructures busi- 
ness," B usiness Week, Ociober 8, 1984b, pp. 124-125. 

[15] LeaviCt, H. J., & Whisler, T. L. Management in the 1980s. Har- 
vard Business Review , 1958, 36, 41-48. 

[16] Leontief, W. Cited in, "Automation's impact on employment," 
Boardroom Reports , February 15, 1985, p. 9. 

[17J Lewis, M. Computer crime: Theft in bits and bytes. Nation' s 
Busines s, 1985, 73(1), 57-58. 

[18] McArthur, D. W. The integrated office: Its present and future. 
Telecommunications , 1984, 18(9), 57-58, 90. 

[19] Memo — A news digest for managers. Office Administration and Auto - 
mation , 1984, 45(7), 22. 

[20] National Bureau of Standards. U. S. Department of Commerce. 

Computer science and technology: Guidance on requirements analy- 
sis for office automation systems . (NBS Special Publication 500- 

















72. C # 80-600179.) Washington, DC: Superintendent of Docu- 
ments, U. S. Government Printing Office, 1980. 

A new era for management. Business Week , April 25, 1983, pp. 50- 
56,60-61, 64, 68-70, 74-76, 80. 

OECD (Organisation for Economic Cooperation and Development). 
Information activities, electronics and telecoomunications tech- 
nologies: Impact on employment, growth and trade , p. 29. Cited 
in The Futurist , December, 1982, p. 37. 

Office automation restructures business. Business Waek , October 
8, 1984, pp. 118-121, 124-126, 130, 134, 138, 142. 
Offices of the future. Newsweek , May 14, 198h, pp. 72-75. 
Pfeffer, J. Power in organizations . Marshfield, MA: Pitman, 
1981, pp. 2-4. 

Picot, A., Klingenberg, H., & Kribzle, H.-P. Office technology: 
A report on attitudes and channel selection from field studies in 
Germany. In M. Eurgoon (ed.), Coianunication Yearbook 6. Beverly 
Hills, CA: SAGE, 1982, pp. 674-692. 

Picot, A. & Rcichwald, R. Office : communication by new text media. 
Manuscript submitted for publication, 1984. 

Pomerdntz, D. Managing information in the eighties. Today's 
Office , 1983, 17(9), 53-56. 

Poppel, H. L. Who needs the office of the future? Harvard 
Business Review , 1982, 60(6), 146-155. 

Porat, M. U. The information economy: Definition and measurement . 
Washington, DC: U. S. Department of Commerce, OT Special Publica- 
tion, 77-12 (1), May, 1977. 

Rice, R. E. & associates (Eds.). The new media: Communication, 
research, and technology . Beverly Hills, CA: SAGE, 1984. 
Rice, R. E. & Bair, J. H. New organizational media and produc- 
tivity. In R. E. Rice and associates (Eds.), The new media: 

Beverly Hills, CA: SAGE, 

Communication, ret'.earch, and technology 
1984, pp. 185-215. 

Rosenberg, A. M. Telemessaging answers the phone 

Age , 1984, 1(7), 30-33, 36-37. 

Spero, J. E. Barriers to international information flows. 
Te 1 ecommunicat ions , 1983, 17(11), 67-69. 

Steers, R. M. Introduction to organizational behavior . Glenview, 
IL: Scott, Foresman, 1984. 

Surviving the paper tide: Xerox integrates information manage- 
ment. A study conducted by Xerox Corporation. El Segundo, CA: 
Xerox, no date, but circulated in 1984. 
Szyperski, N. Computer-conferencing — Einsatzformen und 
organisatorische Auswirkungen. In 0. Grlln & J. Rtiss 1 (Eds.), 
ComputergestUzte Textverarbeitung. Munich: R. Oldenbourg, 1979. 
Uhlig, R., Farber, D. & Bair, J, The office of the future: 
Communications and computers . New York: North-Holland, 1979. 
Valle, J. Computer message systems . New York: McGraw-Hill, 

Vaughn, J. Here comes the mail room! Te lecoimnunications , 1984, 
18(9), 64n-64p. 
Vernon, R. E. The ideal inregrated office. Today's O ffice , 1984, 




19(2), 27-28. 

[42] Walton, R. E. i Vittori, W. New information technology: Organi- 
zational piohlem or opportunity? Office: Technology and people , 
1983, 1, 249-273. 

K3) Welson, L. F. An evolutiona-y move to the office of the future. 
Infosystems , 1979, 6, 78-80. 

[44j What next? A world of communication wonders. U. S. News & Worl i 
R eport , April 9, 1984, pp. 59-62. 

[45 J White-collar jobs: A hot market cools. Business Week , 1984, 2872, 

[46] Witte, E. Die Bedeutung neu. r Kommunikationssys^eme fUr die 

Wil lensbildung im Unternehmen. In A. Alback & D. Sadowski (Eds.), 

Die Bedeutun g gesel Ischaftlicher Vertoderungen fUr 

Wil lensbildung in Unternehmen. Schriften des Vereins fUr 

Sozialpolitik, Heue Folge, Bd. 88. Berlin: Duncker & Humblot, 


[47] Witte, E. Kommunitiationstechnologie. In £. Grochla (Ed.), 

HandwBrterbuch der Organisation . Stuttgart: C. E. Poeschel, 1980. 

[48] Witte, E. Organisatorische Wirkungr.a neuer Kommunikationssysteme. 
Zeitschrift fUr Organisation , 1977, 46. 361-367. 

[49] Women at work. Busines s Week , 1985, 26J6, 80-8S. 

[50] Wynne, B. & Otway, H. J. Information technology, power and 
managers. Office: Technology £nd people , 1983, 2, 43-56. 

[51] The Yankee Group. Report on electronic mail . Boston, MA: The 
Yankee Group, 1978. 







Robert M. Ranftl, Hughes Aircraft Company 

Hughes Aircraft's 13-year productivity study clearly identifies 
management as the key link in the entire productivity chain. This fact led to 
the establishment of a long-term series of seminars on personal, managerial, 
organizational, and operational productivity for all levels and sectors of line 
and staff management. 

To inspire the work force to higher levels of productivity and 
creativity management, itself, must fir^t be inspired. In turn they have to 
clearly understand the productive and creative processes, fashion an effective 
productivity improvement plan with sound strategy and implementation, 
create an optimal environmental chemistry, and provide the outstanding 
leadership necesszu-y to propyl their organizations to achieve full potential. 

The primary goals of the seminars are to (1) ignite that spark of 
inspiration, enabling productive action to follow, (Z) provide participants a 
credible roadmap and effective tools for implementation, and (3) develop a 
dedicated commitment to leadership and productivity ihroughout the 
management team. 


Hughes Aircraft ~ a high-technology organization of some 77,000 
employees'-jnitiated a study in 1973 on means of optimizing productivity in 
technology-based organizations (See Box). The results of this continuing 
study, now in its thirteenth year, clearly stress that management, itself, is 
the key link in the entire productivity chain. 


Skilled, responsible management and superior productivity are 
inseparable; we are entering a far more demanding era requiring greater 

Copyright © 1985, Robert M. Ranftl, P.O. Box 49892, Los Ar.gelss, CA 
90049. Portions of this article include excerpts from R&D IVoducti'tity © 
1974, 78 and other articles previously copyrighted by the author. 

Permissicn to reprint granted by Robert M. Ranftl. 




professionalism ^mong managers. Tomorrow's manager, in addition to being 
technicaly qualified in his or her field, must be a respected, people-oriented 
leader skilled in tt;e latest techniques of behavioral science and sound 
business practice. In tue past, many managers were able to "get by" on their 
technical expertise alone. This, however, will not be possible in the future. 

The critical tie between an organization's management and the 
organization's productivity is evident in the definition of productivity, itself. 
Basically, productivity is the ratio of valuable output to input, i.e., the 
efficiency and effectiveness with which available resources — personnel, 
machines, materials, capital, facilities, energy, and time — are utilized to 
achieve a valuable output. 

Virtually anyone could manage if resources were unlimited. 
However, as we are all well aware, this is seldom, if ever, the case amd, 
therefore, the chsdlenge of creative msmagemeni is to get the job done 
optimally with the available resources. And, looking forward in time, there 
will very likely be fewer rather than a greater abundacce of resources at 
management's disposal, thus creating an even greater challenge. 

Inherently, all resources are bipolar, i.e., they can be fully engaged 
and productively utilized, or, just as readily, they can be underutilized, 
permitted to lie fallow, or counterproductively abused. The "bottom line" of 
an organization's endeavors depends primarily upon the effectiveness with 
which management deploys the resources at its disposal. Management's 
primary responsibility— its fundamental reason for existence — is, and always 
has been, the deployment or "stewardship" of available resources. Thus, 
management is clearly the key link in the entire productivity chain. 

Further confirmation of management's critical tie to productivity is 
evident when one considers personal productivity. The Hughes study showed 
that personal productivity does not correlate significantly with such factors 
as IQ, excellence of education, schools attended, curricula pursued, grades 
achieved, or specialized courses taken since graduation. These factors are 
extremely important since they indicate a person's qualifications, aptitude, 
and potential to perform, i.e., they represent one's credentials. Therefore, 
such factors are of great significance when hiring someone into the 
organization. However, study participemts consistently pointed out that 
among qualified individuals, differences in productivity primarily depend upon 
two key factors, namely, attitude and motivation — first and foremost, the 
attitude and motivation of management, and that, in turn, reflected 
downward euid coupled with the attitude and motivation of the work force. 

To achieve high productivity, it is particularly important that every 
member of management be highly skilled, positively motivated, and totally 
committed. Correspondingly, the same posture is necessary relative to the 
entire work force. But, it must be remembered that the psychological work 
environment is a critical factor in this regard, and it is management who 
establishes the psychological work environment, (See Figure 1 regarding an 
optimal psychological work environment.) 

It is management who gives the challenging assignments or lack 
thereof; it is management who establishes the equity (fairness) within the 


'■*r5^- ■■ . 



organization or lack thereof; it is management who exhibits genuine interest, 
encouragement, and appreciation or lack thereof, maintains equitable 
incentives and rewards or lack thereof, etc., etc. As can be readily seen, 
management not only directly determines its own attitude, motivation, and 
productivity, but through its managerial style and technique, and the 
associated psychological work environment, is extremely catalytic in 
influencing the attitude and motivation — and therefore the productivity — of 
the entire work force. (See Figure 2 regarding Productive Managerial Style 
and Technique.) 

Still further confirmation that management is the key driving factor 
relative to an organization's productivity is evident in another study finding 
which showed that the overall productivity of an organization depends heavily 
upon its management personnel and the top five percent of the staff, i.e., 
people who deal largely in the realm of creative and innovative ideas, 
judgment, major decisions, and actions. Participants did not diminish the 
importance of high productivity on the part of everyone else in the work 
force, but the point they clearly made is that it is the managerial personnel 
and the top five percent of the staff who set the pace for productive 
operations, i.e., it is their ideas, judgment, direction, actions, example, etc. 
that set the pace for organizational productivity down the line. 

If space permitted, many additional proofs that management is the 
key link in the entire productivity chain could be provided. Suffice it to ^ay, 
management definitely is that link. 


Of all factors, leadership has by tar the greatest leverage on 
productivity. Ultimately, the destiny of any organization hinges on the 
caliber of its leadership. Although many managers have distinct leadership 
abilities in certain aspects of their jobs, very few qualify as outstanding 
leaders. The few who do are unusually competent, dynamic, confident 
individuals who somehow "have it all together.' They are the "uncommon 
leaders" potentially capable of creating a renaissance within their respective 

Of significance, leadership is entirely the outgrowth of 
self-deveiopment. No two leadership styles are the same — each style is and 
should remain unique to each individual. Furthermore, a good leader in one 
situation may not be a good leader in a different situation. Also, the type of 
leader needed depends specifically on the group to be led. Yet, even the 
same group may require a different kind of leadership at different times in its 

Fuitdamentally, leadership and management are uniquely different 
entities, with the intrinsic value of "outstanding leadership" orders of 
magnitude above and beyond "effective management." Many individuals score 
satisfactorily as managers who run orderly, profitable organizations and stay 
on par with the competition. The key question is whether those organizations 
are operating anywhere near th-ir full potential. Unfortunately, in the vast 
majority of instances, they are not and very likely neither are the 
organizations with which they are competing. 


-./TPi-^, ,. 

'r«k^^^!4.:^.U -. ^-. . ' H" 

That broad gap separating status quo from full potential can only be 
bridged by outstanding leadership. Such leaders and th*i organizations they 
manage, compete with themselves; they do not permit others to pace them. 
In competing with themselves, they set thsir stssdsrds particularly high, and 
then drive themselves to achieve and transcend those standards. Neither do 
they permit the environment to box them in. What to others may appear to 
be adversity, uncommon leaders look at as challenge and opportunity to 
change and build. Although realists, they have a strong, positive, 
life-enhancing bias. 

Rather than being merely goal oriented, leaders live by values, 
purpose, and commitments, far superior to and transcending finite goals. 
They actively share their values, purpose, and commitments with others, 
stirring others into action. Thus, uncommon leaders create an excitement 
and share that excitement with their followers, permitticg them, also, to rise 
to higher levels. The followers hope some of the benefits of the endeavor will 
rub cff on themselves; this, in turn, happens and synergistically the overall 
effort results in an unpolarized win/win situation wherein all parties involved 
productively come out ahead. 

Leaders not only seek improvement in their organizations; they 
continually seek improvement in themselves. They are highly creative, 
particularly with respect to themselves and their lives; literally, they are 
talented artists who make themselves their own greatest masterpiece. In 
essence, they are on a self-propelled collision course with iheir own destiny. 

Being very people-oriented, leaders value and welcome the abilities 
and potential of their followers. In contrast to fearing and competing with 
their subordinates, leaders seek out and enlist their subordinates' full efforts, 
encouraging them to also expand to their limits. Thus, in an organization 
embodying outstanding leadership, there is minimum rigidity, fear, politics, 
and gamesmanship; intrinsic, not extrinsic motivation is the norm. There is 
both high productivity and a high esprit de corps. 

In summary, leaders bring out the best in people and organizations. 
This is largely because leaden* elicit strong, positive, emotional reactions, 
and people fulfill their needs and grow under effective leadership. Such 
leaders are highly proactive and flexible, focusing on the horizon and beyond, 
continually alert to symptoms, portents of change, and opportimities. Their 
never-ending search is to discover the possibilities and potential in everyone 
and everything encountered. They have an imcanny knack of cutting through 
complexity, providing practical solutions to difficult problems, successfully 
communicating these solutions to others, and instilling enthusiasm and a 
"can-do" attitude. Through such leadership, organizations are dynamically 
transformed — tnere is a strong sense of mission and excitement with people 
and operations propelled to peak performance. (See Figure 3 for the Profile 
of an Outstanding Leader.) 


Hughes Aircraft Company is taking an aggressive, multifaceted, 
systems approach to achieving full potential; focus is on total quality, with 


"- • ''i^:.*'^- - '■' 

'■'?'»;iv v? 


primary attention to product quality. The company believei that to produce 
a high quality product, emphasis on quality must be inherent in every facet of 
the organization and every segment of the product's evolution, i.e., quality 
must be foremost in the mind of every manager euid employee — it must be 
designed, planned, and built into the product. 

The company feels that, collectit'ely, its management and employees 
are by far its most important resource in achieving full potential; therefore, 
particularly great emphasis is being placed on this facet. Other important 
facets include focus on effective marketing and contracting practices. IR&D, 
optimization of design processes, value engineering, product effectiveness, 
improved manufacturing technology, integrated CAD/CAM, effective 
procurement practices, employee involvement programs, performance and 
cost-improvement systems, upgrade of facilities and equipment, etc. 

The company's overall operations improvement effort is 
orchestrated by a corporate-chaired council membered by senior 
representatives from each of the major operating activities throughout the 
company. The scope of this article, however, is not to encompass the overall 
improvement efforts within Hughes, but to focus on a particularly critical 
element, managerial productivity. 


Hughes recognizes that the key element in achieving and sustaining 
high productivity and product excellence is its management team. Therefore, 
it is striving to develop the most productivity-conscious and productive 
management team possible. To help accomplish this objective, the position of 
Corporate Director of Managerial Productivity was created. Basically, the 
charter of this position is to develop the highest caliber, totally committed, 
productivity- and quality-conscious management team possible. The primary 
vehicle in achieving this end is training. 

PRODUCTrvrry courses 

The initial productivity training thrust at Hughes was a long-term 
series of voluntary, after-hours courses for line and staff memagers. Four 
basic courses were provided, comprised of 25 class-hours each. The foin* 
courses focused respectively on (1) personal productivity, (2) manageriad 
productivity, (3) organizational and operational productivity, and (4) effective 
means of combating counterproductivity. A particularly valuable adjunct to 
these courses was that either the Chief Executive Officer or President of 
Hughes Aircraft Company personadly joined each of the classes for one of the 
ten evening sessions. 

After personally teaching more than 1,000 hours of such classes, the 
author recognized that a more streamlined approach would be necessary in 
order to reach all the management of a company comprising 77,000 
employees. Furthermore, although the classes were always fully subscribed, 
it was evident that those managers already predisposed to productivity 
improvement were more likely to volunteer than those most needing the 




A transition was, therefore, made from voluntary, after-hours, 
productivity courses to a long-term series of mandatory, off-site seminars on 
personal, managerial, organiz&tionad, and operational productivity for all 
levels and sectors of line and staff management. The content of these tight, 
fast-paced, one-day seminars, which are conducted during normal working 
hours, is a distillation of the 100 class-hours of material contained in the four 
original productivity courses. The content of the seminars is additive to, and 
complements the material contained in the study report, R&D Productivity; 
the author strives to minimize redundancy realizing that seminar participants 
have either read the report or can do eo at their convenience. 

The seminars ar« not intended to constitute a "cookbook" approach 
to productivity improvement, nor is there any attempt to cast participants 
into a common mold. Quite to the contrary, at the start of each seminar, the 
author points out that managers should develop and maintain their own 
personal style of management, and particularly their own unique style of 
leadership, adding and subtracting to that style only as they, themselves, see 
fit. Therefore, th& seminars do not present a productivity gospel; rather, 
they provide a sharing of insights and ideas. The intention is to (1) develop a 
dedicated commitment to leadership and productivity throughout the 
management team, (2) provide team members a credible roadmap and useful 
tools for implementation, aind (3) stimulate team members to take effective 
productivity improvement actions with respect to themselves and the 
organizations they manage. 

Productivity improvement cannot be dictated or legislated from the 
top. Those at the head of organizations can be highly productive themselves, 
set noteworthy example, and be catalytic in enhancing productivity down the 
line. However, peak organizational productivity will only be achieved when 
each manager stimulates productivity improvement within his or her 
respective organizational sector, with all subordinates performing 
productively in their incumbent positions. 

To date, the author has conducted more than 100 seminars within 
Hughes involving Leveral thousand members of company line and staff 
management. As described below, a typical seminar (8:30AM- 4;30PM) is 
comprised of four modules of approximately equal length. 

Mcdi'ie #1 

The first seminar module establishes an awareness, understanding, 
and com.non vocabulary among the participants. It deads with the anatomy of 
productivity, i.e., what is it, why is it crtically important, what are the key 
factors that impact it, how can it be evaluated, and, of greatest importance 
on an overall systems basis, how can it be improved. 

Module #2 

The second seminar module focuses on managerial, organizational, 
and operational productivity. A number of key counterproductive factors 
common within organizations are identified and analyzed, with focus entirely 




on effective managerial techniques of precluding or combating such 
counterproductivity. Of the many roles and responsibilities of management 
probed in this module, particularly great emphasis is placed on leadership and 
means of enhancing it. - 

Module #3 

The third seminar module focuses on personal productivity. A 
number of counterproductive factors commonly experienced in one's personi>U 
and professional life are identified, with the entire emphasis on effective 
means of precluding or combating such counterproductivity. In conducting 
this portion of the seminar, the author draws upon more than 30 years of 
research on productivity, leadership, management, motivation, creativity, and 
professional self-development. Much of this research focuses or the diaries, 
notebooks, and personal journals of uncommon leaders and creative giants 
who have withstood the test of time and uniquely stand out in history. Such 
materials, never intended for the eyes of others, clearly reveal the core of 
motivation and creativity that drove these unique individuals to the pinnacles. 
The author identifies the many common denominators among these individuals 
and synthesizes these factors into an action plan which one can follow in his 
or her pursuit of professional self-development and excellence. Because 
personal productivity and creativity totally complement each other, and since 
creativity is a key ingredient in leadership., much attention in this module is 
focused on personal creativity and means of enhancing it. 

Module *4 

The fourth seminar module comprises a participative workshop on 
identifying an organization's major sources of counterproductivity and 
formulating constructive means of combating such counterproductivity. The 
workshops ai*e completely positive in nature, based on concensus and 
constructive discontent. The participants recognize that there are already 
many hjghly productive factors within the organization (these also are 
identified); the objective is to make the organization even better. 

The workshops have proven very effective within Hughes and the 
many other organizations for which the author hats conducted seminars on a 
consulting basis. One might assume that most organizations' lists of major 
counterproductive factors would be vexy long; however, the opposite tends to 
be the norm. Nevertheless, those lists, although relatively short, are 
extremely important and should be judiciously dealt with in the productivity 
improvement process. 


Participants' reaction i.o the seminars has been particularly 
favorable; en the anonymous evaluations conducted upon completion, 
participants consistently rate the seminar highly. Utilizing a scale of 1-6, the 
overall evaluation of the several thousand seminar participants within 
Hughes, to date, averages between 5 and 6; the ratings of seminars conducted 
for outside organizations is comparable. Of particular significance, 97 
percent of all the managers who have participated in the productivity 




seminars within Hughes and the many other industrial, commercial, 
governmental, and academic organizations for which the author has 
conducted seminars have recommended that other members of their 
organization's management participate in subsequent seminars. Through such 
endorsement, a domino effect is being achieved which is very important when 
striving to make productivity improvement an integral part of the daily way 
of life throughout an entire organization. 

Further confirmation of the seminars' value is reflected in the 
thousands of written comments submitted by participants. Typical are: 

"Best seminar that I have attended; really helpful, all 
people in management should attend." 

"A must for anyone who has a desire to become a leader." 

"I've applied some of the principles in my daily routine and 
had the g>*eatest results with excellent feedback. " 

"This seminar was very successful in motivating me. I 
believe my increased motivation will be contagious to the 
people working for me." 

"So much infonoation, so well done — a wealth of ideas to 
integrate into a productivity improvement effort." 

"Reawakened my desire to motivate my people to increase 
our total effectiveness and enjoy the fun of doing it." 

"Exceptionally strong reaction — energizing and motivating 
— I feel as if I will go out and go, go, go !li 

Recently, an extensive sample survey of past seminar participants, 
both within Hughes and outside, was conducted. to determine the long-term, 
lasting value of the seminjirs. Although considerable time had elapsed since 
they participated (several months to several years), the large majority of 
respondents (more than 80 percent) still look back on the seminars as being 
highly effective, and feel their personal productivity and that of the 
organizations they manage has been, and continues to be, enhanced as a result 
of the seminar. 


The approach to productivity improvement must be totally 
professional, avoiding all gamesmanship, fads, and buzz words. A well 
qualified, highly motivated and dedicated mamagement team ~ an 
enlightened, people-oriented management team --that optimally deploys the 
organization's resources, particularly the human resources, is the ultimate 
foundation upon which to productively and creatively build. 


The productivity seminars — which are continually evolving and 
g — are making a valuable contribution in this regard. To inspire the 



^v S 


work force to higher levels of prckiuctivity and creativity management, itself, 
must first be inspired. The primary goal of the seminars is to spark that 
inspiration, enabling dedicated commitment and productive action to follow. 

An economic renaissance is critically needed. The effective 
chemistry for such a renaissance requires a synergistic blending of 
people-oriented leadership and a stimulating, creative climate. Only under 
such conditions can — and will — people put forth their best efforts. 
Management's most important responsibility is to establish the climate 
necessary for this renaissance and provide the uncommon leadership to "put it 
all together." 


■t 318 


"^ - •'■^^'••^ ■ 


^ 'Wr-YnX-VSt-V 




This continuing productivity study, now in its thirteenth year, encompasses not only the 
traditional research and development functions, but all the key interfacing activities; 
e.g., marketing, contracts, finance, procurement, manufacturing, information systems, 
human resources, irulustrial relations, and support services. To date, the study has 
involved the active participaiion of 5S major organizations in industry, government, and 
education; the services of 28 prominent consultants; surveys of more thon 3,500 
managers, and an extensive literature search. In addition, particularly valuable source 
material has been derived from the candid insights of many thousands of managers who 
have participated in the author's productivity courses and seminars. 

A book-type study report, R&D Productivity, was published in 1974 and a second 
edition in 1978. Although the report focuses on technology-based organizations, the 
vast majority of its findings are applicable to all types cf organizations, and, therefore, 
the report is used as widely in the commercial, ec-vemmental, and academic sectors as it 
is within industry. Robert M. (Bob) Ranft! has dirscted the study since its inception and 
authored the study reports. 

To date, more than 20,000 copies of the report have been utilized within Hughes and 
more than 125,000 complimentary copies have been shared with interested industrial, 
commercial, governmental, and academic organizations throughout the world. Some 
20,000 U.S. corporations — including 95 of Fortune's top 100 corporations — have 
acquired the report, as well as some 1,000 foreign companies in 50 different countries. 
A large number of these organizations are using the report in their internal management 
development programs, and more than 1 00 universities have adopted cf for use in their 
curricula. The report is retained by more than 700 libraries throughout the world and has 
been translated into a number of languages. 

Hughes will be ha'jpy to provide complimentary copies of R&D Productivity to interested 
individuals as long as the supply lasts. Requests should be addressed to R.M. RanftI, 
Corporate Director of Managerial Productivity, Hughes Aircraft Company, P.O. Box 
1042, El Segundo, California 90245. 


• ----ift-v 




• Skilled and effective management and leadership with outstanaing people in key 

• Cleariy identified organizational objectives and performance goals. 

• Simple organization structure featuring clear lines of direction. 

• High standards of «>taffing. 

• Simplicity in all operations. 

• A stimulating, open, creative climate where everyone can be oneself. 

• Meaningful, challenging assignments. 

• A high dugree of personal job freedom. 

• Minimum constraints, procedures, and red tape. 

• A prevailing sense of equity in all operations. 

• An absence of fe^ t^olitics, and gamesmanship, and the avoidance of any 
connotation of "insiders" vs. "outsiders." 

* The absence of a caste system, i.e., the differentiation between first-and 
second-class roles within the organization. 

* Equitable, parallel promotion ladders. 

* An equitable system of incentives and rewards. 

* A climate conducive to career planning, wherein job security is directly tied to 

• A prevailing "can do" attitude coupled with *he spirit of "ti^'iking 
into everything." 






• Maintain high performance standards promoting peronnal and product axcellanca. 

«• Make a genuine effort to understand subordinates; know their strengths artd 
weaknesses, their primary sources of nnotivation, their career goals, etc. 

• Effectively integrate the abilities of all individuals within the organization, matching 
them to the jobs for which they are best suited. 

• Delegate authority, decision making, and control as far down ttie organization as 

• Manage by expectations; set high standards and high expectations, arxj erwourage 
subordinates to achieve th9m. 

• Involve subordinates in planning, goal setting, and decisions that affect them. 

• Let employees demonstrate their capabilities and grow professionally; help 
subordinates prepare themselves for jobs to which they aspire. 

• Maintain light pressure on subordinates to produce, keeping them fully, but not 
excessively, loaded with work. (This must be done skillfully — mild pressure 
properly applied can stimulate productivity, but excessive pressure can easily 
become counterproductive.) 

• Avoid (1) treating all tasks as maximum effots or special cases, i.e., "crisis" 
management, (2) rush and overtime followed by aeiay, and (3) surprise and 
unexpected changes. 

• Be available to subordinates through an open-door policy, and make it a point 
occasionally to informally drop by their work places. 

• Provide subordinates feedback on their performance; recognize and reward 
achievement; cite mistakes fairly and tactfully. 

• Make a special effort tr help suboru» >ates who are deficient in certain aspects of 
their jobs. 

• Assure that no facet of the organization or individual gots short-changed or 
overemphasized, representing equitably all subordinates and their work to higher 
management. (Whenever possible, have the person who originated a unique idea or 
did a particularly outstanding job be the one to ivief management.) 

• Be sensitive to factors that cause employee dissatisfaction and frustration; get to 
the root of such factors arid resolve conflicts in a timely manner. 

• Serve as a buffer to protect subordinc^es rrom many of the daily administrative and 
operational frustrations. 

• Maintain an effective flow of two-way communication, keeping employees informed 
of the broader aspects of the organization's operations. 

• Avoid imposing personal standards on subordinates and "oversupervising." 


- 'i^«»^V ■ 

f 1^ 



Sttt • particularly potKivt tiampt* a* a parson. Th« outcterHlIng laadar axhlbrta eharactarlsttct that atamp 
him/har a* a "oaraon of r>oia. " 

Typtcal oc««rvation»: 

• la unuaually compatant 

• Hat quality and (tv'icknaia of mind. 

• It particularly craativc. inrravativr. and r>ontra- 
ditionai ~ a uniqua indiy>f<"»''ii. 

• la hiflhly telfmotiva'ad, aatf-con'idtnt. and 

• Hat extrameV high intagrity. vaioai, and atand- 
ardt - ttandi abova organixatlonal polltica and 

• Hat unutually high motivat " it dadicatsd - 
hat a firm tanta of purpota and commltmant - 
It navar talf-tarving. 

• Hat a ttrong potitiva onantaticn. 

• Oiaplaya total aalf-command. 

• Hm a high laval of daaarvad aalf-rtapcwt anu 
••( -aataam. 

• Accaptt tht rela of laadar «ylt*i approprlata 
humility - arfioya thi rot* «rtd la daariy 
acQaptad m • laadar. 

• la yyllling to <i*otk h«rd«r than ottiar mamtiara of 
tha taam. 

• Hm panlcularlv high vitatlty, atamtna, and 
raaarva anargy. 

• la continually aaarcffing/laarin^/davalopingf 
axpandlng/a votvtng . 

• la a "»*lnoar " 

Takat a dynamic approach to activiiiet Tha outatanding laadar approachaa tatki .lylth varva ar>d anthuai- 
atm. It alwayt ortamad toward improvamant. 

Ty!>ical obtarvttiont: 

• It action-oriantad - h«i> a compalling dnva to 
accomplith arvj achiavt. 

• It quici( to tiia up tha marit of peopla, idaat, 
ar>d opportumtiat 

• Utat a persuativa ptrtonality rathar thrn forca 
of power to gat thingt dona 

• It tanaciout — partavarat in tf>a 'aca of obtta- 
clct -- tiwayt tmt thingi througr to tuccatt- 
ful completion. 

• la alwaya willing to "atand up and ba countad" 
— matiaa nacaaaary daciaiona and d(v- - vyhat 
haa to ba dona, tvtn though auch ar may 
ba unpopular and raauit In advaraa s.niciam. 

• Continually **%k» naw arH battar wayt. 

• It a viaionary — iu unuauaHy iliillad at pr*d>c! 
ing futura tachnoiogical and oparational r>aadt 
and applicationt. 

■ Alwaya aaat naw challangat and law fialdt to 

Bringt out the bait in paopla. Getting people to (4o their beat and to worli together affuctivaly ia a special 
talent of tha outttanding iasdar 

Typical obtervationt 

■ It itrongly people-oriented. 

■ Exhibitt great respect for hurr tn dignity. 

■ It particulaily ikilled in motivat.ona' procaaiet 
and in deding with people. 

■ Hat well defined, meaningful goalt and tuc- 
cettfully intpiret attociatax tc help achieve 

* Has confidence m peop'e and ef iectively com- 
municates that confidence. 

• Bnngi aoout dynamic ilynargiarr within groups. 

• is siimulating and catalytic - iniitlUa anthuai- 
aam — maintains mi aacKing organixational 
climate — communicates a "can-do" attitude 
in all tctiona. 

• Helps subordinates achieve their full potential. 

Damonttrates great tkill in directing day-to-day oparationt. The outttanding leader axhibitt unuaual abinty 
n dynamicnlly directing operations for maximum ratuttt. 

Typical obtarvationi 

■ Conceptually intagratet all facatt of the opera- 

• Hat a ttrong tente of timing and limrtt — accu- 
rately tentet "when" and "how much" in 
each situation. 

• Has an uncanny knack for cutting through 
complexity — effectrvely tortt out irrtlevanciet 
and identifiat tfie real driving faciora - pro- 
videt p.'actical toiutions to difficult problamt 
and tuccetsfully communicates thete t>oluticn.$ 
to othert. 

• Sanaat what might go wrong and davaiopa 
cpntinger>cy plant. 

• Maintai.1t ';ontrol of all tituationa, ptirforming 
with relative ease durir>g tinies of stress. 

■ Displays un "elegant" aimplicity in all actiona. 

Mcprinitd from RtO Produitlvilv Copyright '1 974. 78 Robert M Ranftl. P Box 49892. Loi Angelet, CA 90049 
NOTE Th» "Piofil* of tn Outiiir^ding L*id4r" i( ictutlly built on a (ou.,daiion of two ot>ier profilet, the "Profile of t 
Pro<!uciiv« Employe* " and the "Profile of • ProdiK tiv* Manager". The compoiite can be viewed •• three concentric 
circlet with tht "Profile of • Productive Employe* ' it the c*nt*r. turround*d by the "Profile of • Productive M«nsg*r" 
eiul th* "Profi'i of tn Outliending L*«d*r" circumicribad (round th* other two Niturslly, no t'ngi* individutl coil*);' 
livelv (ncompttict ell Ih* trtiii li(t*d in th* thr** profit*! Rather, the profiles provid* * shopping hit of treiti front 
which on* can Mi*ct i or h*r unigu* puriuit of •*ll-d*v*lopment . Unfortunately, ipec* doe* not permit reprinting of 
the two cor* profile ly c*n, h<>w«v*r. b« found in th* (ludy repon. RAD Productivity. (See Boil 




'' Robert M. (Bob) Ranftl is Corporate Director of Managerial Productivity at 

Hughes Aircraft Company, and is President of Ranftl Enterprises, a 
consulting firm totally dedicated to helping organizations achieve their full 
potential. He has provided consulting services and hundreds of in-house 
productivity seminars to & large number of major organizations, nationally 
V. and abroad. Bob is a graduate of the University of MirhigaUt. a guest 

4 lecturer at many of the nation's leading universities, and a member of both 

\ the White House Conference on Productivity and the Department of 

Defense Human Resource Productivity Task Force. 


P.O. Box 104Z, EI Segucdo, CA 90245, Telephone: (213) 414-7872 

P.O. Box 49892, Loa Angeles, CA 90049, Telephone; (213) 471-1804 





' at 













ncREASED Huxwnyrre in flight mm voice gqmmanddc 



AutooBtlc Speech Recognition tednology has natured to the point v^re 
■ . it can provide a viable neans c£ iiKzeasing productivity ly naturalizing the 

mn-nadiine interface. With e\«r increasing woikloads being placed on 
astrooaits, speech recognition m^ provide an altemati^« o»ns of extern 
controlling tliat would reduce the tadc hirden. Voice rmmandlng, alloHlqg 
"handa-free" operation, can be especially effective during qjerations raqulriog 
i slaultaneous syston control. A fll^t eaqjerhaent is ixider development to denoir 

' strste the operational effectiveness of wice control ty connanding the Space 

^ Shuttle's Closed Circuit Telfivislon (OdV ) systenu This experlaent wQl t^lp 

direct future applicatlonB of wice entry to space operations. 


Speech reccgnitlor has nade its wh> into the consuner maiket throu^ 
the personal conputers, toy TOlcenoontrolled robots, ari aids for the 
handicapped. Other uses Include of floe and factory autooHtlon, sortation 
applications, quality control, inwntoty nanagenent, and spedcer verification- 
Govemnent programs such as Advanced Rotocraf t Technology Integration (ARTI ) 
and Advanced Fighter Technology Integration CAFTI ) have devel q»d the uae of 
Autonatlc Speedi Recognition (A£R) in inJLitaiy aircraft fonn helicopters to 

The application cf ASR devices can facilitate slnultaneous systan 
controlling. This 'Vilti-ta*lng" occurs frequent' - -"n spfioeccaf t operations 
when a single crewnErober nust operate more than o cem. One such 

operation occurs during Space Shittle missions whb.. -iin^la crewperson Is 
tadced with operating both the Remote Manipulator Syst^ C<?B) and the CCIV 
system. The ^IS, or payload hay "ann", requires two hand cootrollers for 
operation, and payload nanlpulation Is a visually danandlng ''.adc. Thus, it 
is imdeslreable to remo>« visual and dexterous attention fron RMS operations 
to utilize the GOV system. 

Integration of Automatic Speech Recognition into spacecraft systems 
is under study at NASA's Johnson Space Center vihere development of Shuttle 
Orblter erfiancements and concept designs of Space Station are being conix;ted. 
To demonstrate the capabilities of an ASR device in a space operations en- 
vixxxctsnt, an experiment will he flown to control the Orblter's OCIV system 
by the Voice Conmai^ Systan (VCS ). 



( i 



Autonatlc Speech Recognld.on (ASR) is the ability to discern and 
dlstli^sh utterances. This is done chiefly ty pattam matching techniques. 
Ewry utterarce tl»t is to be r.\xgnlzed is represented ly t. "patten" that 
dlstiiguislcs It fron all other patterns. A spoken \T)rd can be processed into 
a pattera and coipared to eidsting pattens in nemory. Tljest patterns can be 
crated ty tte user repeating words into an ASR de'/ice (spedcer dependent 
rec jnitlon), or, they can be more generalized, created ly ccnijinliig together 
the pauteros that dlstlngiulsh the basic phonemes or sounds of a glwan dtalert: 
to represent a spcken word (spedcer independent recogtdtlon). Once there are 
patterns representing a set of words to be recognized, an Affi. device ad^ 
reccgnlze words spoken discretely, one-at-a- t l tn? (iaolatad) or ml^t haw the 
capability o£ recognizing words spcken cootlnwusly as in natural speech 
ccnnunlcatloQB (caatiiuaus). 

The most reliable state of the art devices are the spe^^r dependent, 
Isolated word reoognlzers. However, then, are sone dependant contliuous 
recognizers that are hl^y reliable depending on thier application- Ifast 
Iniepeixient speedi recognizers can onl> woA. reliable with snail vocabularies 
and are nnst accurate only tor chosen dialects of a given laiigjage. ASR can 
\y- used as a tool for language understanding which Inwlves the fomridible 
taA. of context interpertation. language understanding is considered an 
artificial intellegenoe tadc. In the fligj't experiment, a state of the art 
dependent recognizer with contliuous recognition capability will be used. 


The Voice Ccomand System experiment package will consist of a OdV 
system Interface, a control panel, and an ASR device. Figure 1 shows the 
placement of the VCS mlt iiit the Orhlter's AFT fll^t deck. 

"ae _»__?□* 

«c( conraOL rtMti. 

rlsur. \ 


31' 6 



For Mnlnum Impart on system Interfacing, the VCS will control the 
GOV systan }y paralleling the switches oo the COV control panel. OCIV 
coBiBDCls InJude: selecting DooltoXB, caaeras, panning, tilting, focusing, 
nxndic aid iris control. These comanda result in visual reactions. That 
is, something happens that can be corflmBd visually; selectir^ a camera will 
show 38 a displav change on a monitor; pamlng vdLl show as monitor dlsplayr 
pandic ani 80 on. Also, the CX3V cortrol panel nas ligjited pusMuttots for 
most camands, so there is a second visual feedback. With this existing feed~ 
beck, there is vo need for a display on the VCS. There vdll be an audible 
feedback for confirming recognition or rejectioa of an Input utterranoe. 

Normal crew audio ooamnlcation is a rsiuireDient of the VCS. This, 
there will be an interface to the Shuttle ^udLo Distribution Systaiu This 
cooBction vdll allow other cronEobexs and groundUik to camxilcate with 
ths VCS operator. A "relax" state of the VCS will allcwthis commlcatlon 
dLsr^id the operstor'3 input until a "reedy" state is enabled. This will 
be acconpUs'^ \3j a panel a^tch on the VCS or ty a voice conoand. 


The objective of this experiment is to demonstrate the ef fectiwness 
of vjice entry to increase the efficiency and productivity of the crwmenber 
while reducing the tajk burden- Since ASR devices are tools to npnipulate end 
ijystenB and are not in of themselves an end system, their effectiveness is 
measured subjectively ty how the qjerator peroei^«s the systan's effectl\eness. 
A technically gixxl speech recognizer may fall because the user did not want 
it to woik or perceived no adxantage to using the device. While speech is 
a natural and ccin\«nlent method of conmntting systems, it is often applied 
unnaturally. For instance, the cocBtraints of an isolated recognizer to say 
4 words one at a time with pauses, may be a hinderanoe lAien controllinc a systan 

that requires a contliuous stream of caoiBiids. As an eiom^Je, the poet office 
ml^x use \rolce recognition to help sort mall. The operator would tike a letter 
or package, read the zip code and let the conputer determine the sorting. It 
would be less efficient to s^ the zip code one mnter at a time rather than 
the full sequence. 

Other factors that influence the acceptability and perfotmanoe of an 
ASR device are: 

o PJysiological o Psychological o Envlrotmental 

o age o mDti\atJ.on o noise 

o sex o attitude o atrooephere 

o fatigue o stress o pressure 

o dialect 
o health 





One udque problem to this application of wlce control is the effect 
of nilcro-graAd.ty on tha hunan speech nechanlam. Poeslblf effects includt>: 

o No pull on tongpe or cUaphEagm 

o Atroply of \Dcal tract muv.lew for long diration exposure 

o Sims oongestloo, vMdi requireB gravity for draining 

o Special atmosphere mixture and pressure 

An atteqpt was made to help discern aof degradation of speech In an ex^'erlment 
floun on the sixth Shuttle mission. Astronaut Sto:y Misgram helped conduct 
an experiment ty making xoice recoidliigs before, during, and after his mission. 
Subsequent testing on these recordings did show a drop in recognition accuracy 
for pre-nission trained patterns on a discrete ASR device. However, only 
the one test subject, who vas fatigipd, and the other error factoos like the 
effects on the tape reooidcr, the results of the experiment are vaaxachjlai^. 
The experlmsnt indicated further testing is required. ReconUqgs are ag^ 
planned as part of the CCIV experiment on magnetic tape as tell as data stoned 
into mBmoiy. 

Recognition results are best on a dependent ASR device v*ien trained in 
the eovironnent it is to he used in» The plyslolqglcal effects encamtered 
during a mission can only be slmilated. The micro-gravity environmeot camot 
be. Houever, t'e anMent noise on the shittle has been measured dirlng fllg^ 
by sound ievel suneys conducted on the first three Space Shittle missions. 
This noise can be recreated for training patterns in a fli^rtrtype environoent. 


Passible future applications for voice iiput/output in a Space Station 
design include: 

o Control of closed circuit television systan 

o Infonnation retleval hy audio resporse 

o SystenB checkout nad \erlfication 

o Data retrieval during EVA 

o Control of robctlcs 

o Inventory nanagement 

o Maintenance and repair aid 

o Systan configuration 

Voice entry can be applied equally well to ground operatlonB with similar 
applications. The technology is evolving to include large vocabularies 
(10,000 words) with epedcer Independency. With the added capability of 
natural language processing, voice l^) may achieve the man-machine comnunica- 
tion ability that reseiribles the hunmr-to-hinuin comuinlcatlon. Yet, such goals 
are still ehiaivi. 





Results from the fUg^ experiment will be neasured from the Voice 
Command System's memoiy to detetoine the accuracy of the ASR In flight and 
from the subjectlw repotting of the cra^nenber/operator. This flret-tlme 
appllGition of wlce reoognltloa to a spacecraft envLnxment will be a fore- 
runner to de\elcpli)g standard voice 1^ for Space Station and future NASA 


WllUam T. Jordan recel\)ed his B.S. In Electrical Engineering from the 
Uni^)erslty of Texas In May 1980. He has since been osplcyed ty the National 
Aeionaitics and Space Adninlstratlon at the Lyndon B. Johnson Space Center 
in Houston as the Project Engineer on Voice Cocnanded S> stems. 






John M. Nicholas, Loyola University of Chicago 
Ronald S. Kagan, LaJolla, California 


Space Station crews will encounter new problems, many derived from 
the social Interaction of groups working in space for extended durations. 
Solutions to these problems must focus on the structure of groups and the 
interaction of individuals. A model of intervention is proposed to address 
problems of interpersonal relationships and emotional stress, and improve 
the morale, cohesiveness, and productivity of astronaut teams. 


With Space Station, the nature of American Ipace missions will 
change. One difference is that NASA will be shifting from missions of 
relatively short duration to operation of a permanent, continuously- 
occupied platform. Astronauts will be living and working in space for 
weeks and months. Interplanetary missions of the future will take years. 

A second difference is that Space Station will increase the size 
of the ongoing crew complement. The preliminary design plan calls for six 
to eight people living and working for 90-day stretches in 300-mile orbit. 
The number of people in space for extended periods will continuously grow 
as Space Station is expanded and others are built. 

Third, the makeup of astronaut crews will be different. For two 
decades, NASA selected highly trained and disciplined military test pilots, 
men who would do as they were told and consistently perform well, even 
under extremely stressful and hazardous conditions. In time, NASA will 
rely more and more on industry and academia for its astronauts. Whatever 
their expertise, one thing is certain: future space crews will be made 
of Different Stuff, different both from the astronauts of yesterday and 
different from each other. They will also devote much less time in prep- 
aration for space flight, so not only will they be different from the 
start, they are likely to stay that way as astronauts. 

*Appreciation is extended to Gustave J. Rath of Northwestern University 
for his helpful comments. 


j^F "- -'V^SSii-..-' 





Finally, more decision making and problem solving will be made 
In space by teams of astronauts themselves . Thus far most work In space 
has been performed according to schedule and checklist, and with the 
assistance of Mission Control. As man stays longer in space, more con- 
tingencies will arise and more judgement will be required by astronauts. 

New Missions, New Problems 

As the size of teams and amount of time they spend in space in- 
crease, so does the likelihood of dysfunctional group behavior Behavior 
of heterogeneous groups of six or more in extended isolation is different 
than that of two- or three-man homogeneous groups, so problems can be ex- 
^ pected that were not seen on earlier U.S. space missions. Informal pat- 

* terns of behavior common to work groups will arise in space. Looking at 

small groups in confined. Isolated, and hazardous conditions on earth and 
aboard Soviet and American space vehicles suggests some of the kinds of 
social /psychological problems expected aboard Space Station. Unfortunately, 
most observations have focused solely on physiological and psychological 
responses; few studies, and none in space, have been directed at group 

Consider, for example, research centered on groups of scientists 
and Naval personnel at remote stations in Antarctica. Among widespread 
4. emotional problems reported during winter isolation are sleep difficulties, 

*. headaches, "feeling blue," feeling lonely, and irritability and annoyance 

„( with others. Studies of group behavior show clear evidence of deteriora- 

tion in social relationships and work effectiveness, particulariLy during 
' " the latter part of confinement. Groups reporting the greatest <iecline in 

teamwork and efficiency also have the greatest persistent difficulty in 
r. keeping essential equipment operating, the most frequent open conflict 

among members, and the lowest morale. 

StrRss symptoms reported in confined. Isolated groups on earth are 
similar to those observed on U.S. and Soviet space flights. CoEimonauts 
report that interpersonal hostility starts to develop after about 30 days 
in space. Hostility also develops between the space crews and mission 
control; cosmonauts report being relieved when communications with earth 
_ were interrupted. Interpersonal difficulties on U.S. flights also produced '* -j 

problems: conflicts on some Apollo missions resulted in faulty cr scrubbed ^ 

experiments; ineffective communication among Apollo crew member; caused ' 

near-fatal mistakes on reentry following the Apollo-Soyuz linkup ll]; and : 

on Sky lab IV, a frustrated, overworked crew went on strike six v/eeks into 
the mission. " 

Space is a high-cost, high-reward environment. Costs include 
overcrowding and daily reminders of physical danger; rewards include 
recognition, pride, and adventure. For groups in hazardous conditions 
i such as space, the costs — being physical stimuli — are relatively stable; 

' » the rewards-mostly subjective — are unstable. Radloff and Helmreich [8] 

"^--"^ believe that "pioneering" groups adjust and perform well because the per- 

/jL-i ceived rewards of the mission exceed the costs, but they predict that as 

T S- time passes, the cost-reward ratio will be altered so that costs exceed 

; * 



I - 

rewards. As space travel becomes more routine, rewards may decline more 

rapidly than the costs. More personnel will be needed, and they will be 

younger and less-experienced. Not all o£ them will view space flight 

with the same sense of accomplishment and adventure as earlier astronauts, 

nor will they be as prepared. Radloff and Helmrelch predict that combined .■% 

changes In the quality of the personnel and In the perceived reward of 

routine work In high-cost situations will result In personnel problems and 

declining teamwork. 

Teamwork Is Important to productivity. Astronauts' ability to 
work together effectively, and to adopt and develop formal and informal 
mechanisms to improve their working relationships and group output is 
vital to the success of Space Station. With the estimated cost of a 
human work hour aboard Space Station being $80,000, there .-.ust be con- 
siderable emphasis on productivity. Yet the stresses on teams living 
and working in space have very high potential for inhibiting communication, 
instigating hostility and conflict, and reducing productivity. 

Current Approaches 

NASA is taking a multifaceted approach to these problems, includ- 
inff aetrcndut selection and training, and human factors design. The U.S. 
and Soviets use psychological/psychiatric guidelines for the selection 
of people with the greatest likelihood of adjusting to the stresses of 
space flight. Both space programs emphasize individual-level solutions 
to reducing and adapting to stress, including relaxation tecnniques, 
meditation, biofeedback, exercise, proper sleep and special diets. 

A second approach is ergonomics and human factors — designing and 
operating a space-habitat that is congenial to living and working in ' 

space. This includes attention to anthropooietric and architectural 
details of interior layout, lighting, textures, color, windows, and 
size and shape of living and working spaces. Private crew-member re- 
treats, control of noise and vibration, programs to encou;:age sleep, and 
effective use of diet and music to combat monotony are examples of features 
being studied. 

Selection, training, and human factors address some but not all « 

parts of the problem. First, despite their obvious value, these approaches j 

ignore group-centered problems and solutions. A common, fallacious assump- [ 

tion is that the process used by effective work teams is natural and even | 

simple. But effective teams do not just happen; great skill and under- i 

standing is required by both the leader and team members. Even groups I 

working under the best possible ondltions on earth suffer from problems ' 

which hinder their ability to perform and limit their output. Secondly, 

current approaches Ignore the social environment and individuals' need 

for "social support." Social support — emotional concern, aid, and 

information from others — is crucial to making social environments less 

stressful, more conducive to stress adaption, and more productive [3]. 

Astronaut crews train intensively under conditions <^s reallsdc as pos- 
sible. They practice respon&es to routine situations and crisis events. Now we 







must ask: What kinds of skills and abilities should they have to make 
them effective Ir. the long-duration, team efforts of Space Station, and 
how should these skills be acquired? 

The technology of applied behavioral science offers interventions 
to increase effectiveness of astronaut teams. This paper proposes an 
intervention model to Insure that space crews, as teams, are functionally 
and social "fit" for missions, and remain that way in space. 


Applied behavioral science in the .^pace Station Program alms to 
improve the Interpersonal effectiveness, stresshandling capability, morale, 
-t' and productivity of crews and their support groups on earth. It includes 

% both proactive and reactive measures — interventions applied in planning 

and preparing teams for space and again during missions whenever problems 

The intervention model is based upon the technology of organization 
development and the use of social support in stress management. Organiza- 
tion development refers to efforts to improve an organization's problem- 
solving and renewal processes through use of applied behavioral science 
theory and techniques (see e.g. [4]). Problem solving is how an organi- 
zation goes about diagnosing problems and making decisions; renewal means 
establishing and encouraging organizational creativity. Innovation, flexi- 
bility, adaptability, and motivation. 

Organization development has six areas of emphasis in the Space 
Station Program: (1) It is preventive work rather than therapeutic; 
(2) it is an advisory /suKEestive approach rather than authoritarian/ 
directive; (3) it places hea"y emphasis on instruction to astronauts 
about relevant aspects of group and Interpersonal behavior; (4) it is 
based on data collection about group performance and '-lehavior; (5) it is 
oriented toward the Space Station crew (as a team) and its problems rather 
than individuals; and (6) it places prime emphasis on development of 
social-support skills and alleviation of social-related stressors, in- 
cluding interpevsonal and Intergroup conflict, perception distortions, 
poor communication, and low levels of trust and openness. 

Behavioral science Interventions focus on the structure of groups 
and the interaction of individuals. Structure Includes the hierarchy and 
formality of reporting relationships, role definition of group members, and 
the degree of flexibility cf the arrangement to accommoda^e the nature 
of decisions and cj^sks. Structure-centered interventions for Space Station 
focus on both space crews and Mission Control. 

Interaction-centered activities, also referred to as process ii.ter- 
ventlons since they concern the process by which groups interact and worK, 
focus on Interpersonal and intergroup communications, Lonflict resolution, 
decision making, and trust and openness among groun members. 




Group Structur* and the Technological Imperative 

Structural interventions have been used to Improve the performance 
of groups by tailoring the group's size and configuration to alleviate 
structural conflicts which Impede work. Research showb that decentralized 
groups are more adaptive to the uncertainties of thiair environment and 
technology. As work becomes more routine, centralized hierarchies are 
more effective. This supports the "technological imperative:" group 
form (its structur'i) follows its function (or work) [9]. 

Group efficiency can be improved by adjusting gro',,-) structural 
(.haracteristics to suit the work. On a recent Shuttle flight, astronauts 
on a two-man team ran into technical difficulties while deploying a 
commercial satellite. After repeated tries, the team was enlarged to 
include a group on earth headed by Sally Ride. Normal communication 
through a central liaison was circumvented as Dr. Ride talked directly 
with the team in space. This new structure, decentralized for Intense 
problem-solving, demonstrates the need for a variety of configurations 
'n space to handle different types of tasks. 

Though the necessity for a shift in structur<> may be recognized 
at the time, such a shift can cause problems with clarity of responsl ~ 
bilities, stability of lines of authority, and morale. The situation 
becomes more difficult as lines of authotlty must rapiuly decentralize 
to allow more creative problem-sclving behavior. Such a change might be 
interpreted as insubordination and be resisted by the flight cocmiander 
or Mission Control. Future mission teams must be equipped with skills 
to Interpret and work with shifting, alternative group structures. 

Group and Individual Process Interventions 

The objective of process-ce'^^tered J'-terventions in the Space 
Station P ogram Is to build cohesive work teams, improve methods of 
resolving conflict^ and develop good interpersonal, communication, and 
emotional support skUIs. Many of the problems and stresses of Space 
Station work inhere in group process — both within and between space crewc 
and terrestrial support crews. The organization development model deals 
directly with these problems in ways which individually-focused programs 

Groups which possess a high degree of clo;^eness and solidarlty> 
and a low degree of tension, hostility, and major conflict are said to b^ 
cohesive. Cohesion may lead to more Investment in the group, Jiore commit- 
ment to group issues, a greater sense of personal security, and the desire 
to be a "good" team memler. These alone seem sufficient for wanting Space 
Station teams to be cohesive, but cohesion is also related to task perform- 
ance. When groups strive for high performance, the highly cohesive groups 
perform better. They are also more adaptable to stressful conditions: 
among military combat units in battle, th" more cohesive units generally 
perform better. 

What can be done to incre^ je a crew's cohesiveness? Two contrlb- 
itlng factors — frequent interaction among members and group prestige — come 


already built-in to space station teams. Cohesion may be further enhanced 
by selecting people more likely to get along — those with similar values, 
personalities, and norms, and evaluating and rewarding them based on group 
performance. Other factors needed for cohesiveness — mutual trust and 
support, and agreement on the thrust and direction of the team — are fre- 
quently absent in groups, but these can be developed and strengthened 
through organization development interventions. 

On the negative side, cohesion may cause members to be more con- 
cerned with the group than with the purpose for which it exists. Members 
choose to ignore the negative aspects of the group cr are unduly influenced 
by other members of the group. This malady has been termed "groupthink;" 
its symptoms and effects have been well-documented [5]. On Space Station, 
this could lead to crews deviating from mission objectives or ignoring 
directives from mission control. Occasionally such autonomy may be neces- 
sary and desirable, but frequent excursions into group-think could result 
in poor decisions or even disaster. These problems are acknowledged and 
dealt with in organization development interventions. 


The intervention model uses a three-phased program of training 
and teambuildiiig, process consultation, and transition debriefing. Action- 
research oriented, the program is directed at maximizing mission-team 
effectiveness and productivity by minimizing tivsfunctional group behavior. 
The intervention techniques, al-hough new to the space program, have shown 
substantial success in a wide variety of settings (see e.g. [2,6,7]). 

Phase 1 • Pre-Fl-ight Training in Group Proces ° and Structure 

il Pre-flight trainin^^ is necessary to provide crews with learning 

* in groap dynamics and structure, leadership, and interpersonal relation- 

^ ships. Some training should be done in unstructured group sessions, where 

; the group :'s allowed to "evolve," sometimes into an informal structure, 

sometimes into disintegration. The trainer encourages individuals to 
explore their feelings, behaviors, the behaviors of others, and the effect 
of th3se on the group. This training can be combined with cognitive 
lectures, structured exercises, and role plays to enhance learning and 
relate it directly to the mission environment. 

To help reduce structural conflicts, crews must also learn con- 
cepts of group-structure analysis, including diagnosis, definition of 
alternatives, evalua«-lon and selection of new structures, and implementa- 
tion and refinement. During training, crews learn to diagnose group 
structure, using responsibility charts and network analysis to highlight 
r communication flows, subgroups, and power and control issues. Crews 

I . ientify the most suitable configuration" for each type of task, such 

" as subgroups, liaison arrangements, or "group-leveling" for non-mission 

activities. Each type of structure is well-defined during training so 
as t.i illicit a set of clear behavioral expectations during missions. 
Selection of a structural arrangement should be endorsed by the leader 
and have the commitment of the group. Crews would learn how to walk 




•-^ ,1,'. I k V\ ■■(. >v» .-S^ 


through the pros and cons of alternative arrangements. Even though roles 
and structures should be well-defined prior to a mission, crews should 
be able to quickly evaluate and refine the group structure to best suit 
unplanned situations. 

Group training also provides skills needed for astronauts to 
support one another in space. Among possible sources of social support — 
work, nonwork, and professional, research shows work-related sources are 
the most effective in reducing work-related stress 13]. In space, fellow 
astronauts are at once more accessible, more familiar with and similar 
in their experience, and more attuned to their unique problems than any 
external person can be. Nonwork sources of support such as interactive 
television meetings with family and friends should be counted as contrib- 
utory stress buffers, but not a major factor in buffering or compensating 
for the effects of work-related stress in space. Few Americars "naturally" 
possess good support skills; the competitiveness and task-orientedness 
of our society generally limit their development. Although emotional 
support is difficult to "teach" as a skill, group training provides the 
necessary experience. 

Interpersonal and group skills in space will become more important 
p«? mankind spends more time there. Notes Cosmonaut Valery Ryumln, veteran 
of two six-month flights, "We (the crew) must solve our problems together, 
taking into account the feelings of the other. Here we are totally 
alone. One must bear in mind constantly the other's good and bad sides, 
anticipate his thinking and the ramifications of a wrong utterance blown 
out of proportion." [1] 

Pre-Flight Teambuilding 

Teambuilding is a proactive intervention to "develop" Space 
Station crews and group support units into effective work teams and 
reduce the likelihood of many group problems ever occurring. Teambuilding 
helps groups to maximize use of members' resources, develop a high level 
of motivation, reduce hostility and conflict among members, and overcome 
problems of apathy — including loss of productivity and innovatlveness. 

Typical teambuilding uses data collection, group workshops, and 
a consultant to help the group analyze its behavior, diagnose its problems, 
and develop plans for becoming more effective. Data is gathered from 
th» ream through questionnaires, interviews, or process observation about 
group tasks, group process, and interpersonal conflict. During the work- 
shop, the consultant helps the group understand and diagnose itb g'-cup 
process. The group develops a plan for becoming more effective and then 
works to implement the plan. 

Traditional teambuilding must be modified for Space Station since 
crews will not be stable. Some astronauts will be on board only once, 
others will be on rotation. The concept of a "Space Station team" must 
be altered, perhaps to include all astronauts to be sent aloft in the 
forseeable future and key figures at Mission Control who will be in 
frequent communication with them. NASA astronauts would be involved 
in teambuilding sessions as part of their initial training. Then — perhaps 


V .' 

every six months — teambuildlng sessions would be conducted to Include 
everyone scheduled to be oi Space Station and at Mission Control for 
the following six months. 

The likelihood of "too-cohesive" groups developing Is lessened by 
awareness of its symptoms and effects. Pre-f light training provides 
astronauts and support personnel with skills to recognize and deal with 
these behaviors, and teambuildlng is canducted so that any one "team" 
is not pitted against other groups with which it shares goals and must 

Phase 2; In-F l' 'ght Process Observation and Consultation 

Process observation uses a skilled third party who observes the 
group and "feeds ' data back to the group. She does not offer "expert" 
advice, but "intervenes" to help the group use its own resources to 
Identify and solve problems Involving the task, the process used to accom- 
plish the task, and interpersonal conflict. The process consultant may 
be anyone trained in process observation and consultation, including 
astronauts or former astronauts. 

The process consultant is interested in the nature and style of 
communication, both overt and covert, the functional roles of group members, 
the extent to which personal needs are shared, activities directed toward 
holding the group together as a cohesive team, problem solving and decision 
making activities, the group's understanding and articulating group 
norms — especially the dysfunctional ones, and the group's understanding 
and coping with different leadership styles. 

A full-time process consultant is unlikely to be onboard to observe 
groups and provide consultation. One alternative is to monitor crews via 
radio or television, or collect data from videotapes, televised personal 
Interviews or computer-assisted questionnaires. She would then provide 
feedback and counseling, in confidentiality, to individual astronauts 
or entire crews while in space to help them become proficient In self- 
monitoring and self-correcting their behavior. The feedback signal could 
be scrambled to insure confidentiality. 

How effective will "remote" process observers be? This is diffi- 
cult to foresee since traditionally the process observer is present with 
the group. But the time, budgetary, and physical constraints of space 
flight will require innovation and adaptation for uany interventions, 
including process observation. 

Alternatively, one of the astronauts onboard could serve as process 
consultant. This would be a NASA crew member with other responsibilities — 
technician, engineer, mission specialist — but who has been trained and 
has skills in process observation and in giving nonevaluative feedback. 

Groups must be familiar with process consultation before it can 
work. The pre-flight training in Phase 1 is good preparation. Ideally, 
everyone in t^e crew becomes a process consultant. 


Phase 3: Post-Flight Transition Debriefing 

A transition meeting is held at the end of a mission or when crew 
members are rotated back to earth. Meetings held onboard Space Station 
would serve several objectives: departing crew would (1) give emotional 
and performance feedback to associates remaining on the station, (2) obtain 
feedback about their performance, and (3) inform new crew members of vari- 
ous aspects of the mission. The procedure is aimed at reducing the 
temporary negative effect of Introducing "novitiates" into previously 
established teams. 

Once back on earth, rotated members would meet with support person- 
nel from Mission Control for a feedback session, the intent being to analyze 
and correct any unresolved conflicts or Issues, and reduce the likelihood 
of similar problems occurring on future flights. The process consultant 
for the mission should be present to help the group identify (or recall) 
and work through communication, interpersonal, or conflict Issues. Again, 
process training and members' understanding of process-related issues is 
a prerequisite for this work. 



A model of interventions based upon organization development and 
social support is proposed to improve team morale, coheslveness and 
productivity, and reduce problems of social Interaction and emotional 
stress. The model uses training, teambuildlng, process consultation, 
and transition debriefing in a three-phased program. Although new to 
the space effort, similar Interventions have been substantially success- 
ful in other settings. The arguments and evidence in favor of the model 
are compelling, while the costs are relatively low and entail no major 
risks. Implementing the model does not require elaborate equipment nor 
take substantial time away from work activity. Initial efforts must 
necessarily be innovative and somewhat experimental, and involve col- 
laboration among behavioral scientists, astronauts, and support personnel 
at Mission Control. 


[1] Bluth, B.J., "Pilots of Outer Space." Society , (Jan/Feb. 1984), 
pp. 31-36. 

[2] Golembiewski, R.T., C.W. Proehl, & D. Sink, "Estimating the 

Success of OD Applications," Training and Development Journal , 
(April 1982), pp. 86-95. 

[3] House, J.S., Work Stress and Social Support , Addlson-Wesley , 1981. 

[4] Huse, E.F., & T.G. Cummings, Organization Development and Change ^ 
3rd ed.. West Publishing, 1985. 


A'-.! ,''»t,ilk.»;;>'-:5^. 


[5] Janus, I.L., Victims of Groupthlnk , Houghton-Mifflin, 1972. 

[6] Nicholas, J.M. , "The Comparative Impact of Organization Development 
Interventions on Hard Criteria Measures," Academy of Management 
Review , Vol. 7, No. 4 (Oct. 1982), pp. 531-542. 

[7] Porras, J. I., & P.O. Berg, "The Impact of Organization Development," 

Academy of Management Review , Vol. 3, No. 2 (April, 1978), pp. 249-266. 

[8] Radloff, R., & R. Helmreich, Groups Under Stress; Psychological 
Research in SEALAB II , Appleton-Century-Crof ts, 1968. 

[9] Udy, S.H., "'Bureaucracy' and 'Rationality' in Weber's Organization 
Theory: An Empirical Study," American Sociological Review , 
Vol. 23, No. 6 (December, 1959), pp. 791-795. 

JOHN NICHOLAS is an associate professor of Management Science 
at Loyola University of Chicago and a part-time management consultant. 
He has a combined background in social and behavioral systems, operations 
research, and aerospace engineering. RON KAGAN is president of Horizon 
Office Systems, LaJolla, California, a consulting firm specializing in 
the use and acquisition of technology in organizations. His experience 
includes organization development, computing science, and physics. 


3 _ t) 

•♦X.^^;^'^%^^■^ --^^ '-■ ■ il, 






SEPTEKKB 10-11, 1985 


A Driving Factor 1r» Space Station Oesign 

Harry L. Wolbers, Ph.D. 

Manager - Man-Machine Systems 

Space Station Program 

McDonnell Douglas Astronautics Company 

5301 Bolsa Avenue 

Huntington Beach, California 92647 

(714) 896-4754 




- A Driving Factor 1n Space Station Design - 

Harry L. Wolbers, McDonnell Douglas Astronautics Company 


The c 'terla of performance , cost , and mission success probabil - 
ity (program confidence) are the principal factors that program or pro- 
ject managers and system engineers use In selecting the optimum design 
approach for meeting mission objectives. A frame of reference Is dis- 
cussed In which the Interrelationships of these pertinent parameters 
can be made visible, and from which rational or Informed decisions can 
be derived regarding the potential Impact of adjustments 1n crew pro- 
ductivity on total Space Station System effectiveness. 



Crew productivity will be the driving factor In determining the 
cost effectiveness of the Space Station and cost effectiveness In turn 
will be the <ey to providing a viable national resource that can meet 
the evolving needs of multiple users during the coming decades. 

\ The very Importance, however, of the problem of maximizing the 

productivity of the crew within the assigned resources and anticipated 
operational constraints of the Space Station emphasizes the need to 
understand the nature and effects of adjustments In crew productivity 
within the overall operational context of the Space Station Program. 

The subtleties of human-machine Interactions dc not always reflect the 
obvious. Conventional wisdom, for example, suggests that: 


f 1. Productivity measures of the crew directly reflect changes 

i In the total efficiency of the (spacel operation. 


-\ 2. Changes In product1v1cy can be measured by output per man- 

*l hour. 


I 3. Increases In output per man-hour are desirable because they 

I yield decreases In user charges. 

I These expressions of "conventional wisdom" are not necessarily 

true. Increasing the productivity of the crew In a given task may re- 


V^^- ^ ■■• ^ \^ 

quire more electrical power, and that 1n turn may decrease the re- 
sources available to support critical payloads and thus degrade the 
total efficiency of the overall space operation. Increasing the output 
per man-hour may cause more errors or mistakes In critical operations, 
leading to poorer overall system performance and higher costs to the 
potential customers. On the other hand, depending upon corollary fac- 
tors, either Increasing or decreasing crew work load can have a nega- 
tive Impact on productivity In view of the long-term effects of work 
values on crew morale and psychological well-being. It must be recog- 
nized at the outset that for the Space Station designer, productivity 
adjustments can be appraised meaningfully oioy within a total system 
engineering framework. The Space Station designer or system engineer 
must consider all the Interrelationships of all the human and mocMne 
elements that are In olved In achieving the mission goals and object- 

In order to analyze the Impact of crew productivity adjustments 
1n terms that will be useful to the system designer, organizations and 
Individuals concerned wUh enhancing space crew productivity must: 

Identify potential ways to enhance overall system productiv- 

Develop criteria to measure changes In productivity In terms 
of overall system performance, cost factors, and mission 
success factors. 

Provide Insight to the system designers of the successive 
linkages whereby crew productivity adjustments can affect 
total system performance, cost, and the probability of 
achieving a successful mission. 

The following paragraphs discuss each of these three key steps. 

Factors that can affect productivity Include both operational 
Issues (such as physical conditioning exercises, training, maintenance, 
logistics scheduling of activities, organization, etc.), and design 
factors (such as Interior layouts, modularity, lighting, noise control, 
work station design, etc.). 

The methodology used 1n defining the critical design and opera- 
tional Issues should be predicated upon a detailed consideration of thp 
classes of crew activities that will be required In the Space Station 
era. These Include three key areas: habitation and survival activit- 
ies; mission and payload-orlented activities; and STS/Ground Interface 
activities. Identification of these activities provides a basis for 
defining operational Issues affecting crew productivity In each activ- 
ity area and a basis for assessing the productivity adjustments that 
could result from design changes. 


• ^"'rsj-^ 

^<^N^-.^i5^; !•' .\> .':■--.-■>., . ^ *-/ 

An example of an operational Issue might be the time to be de- 
it voted to exercise using a particular protocol to maintain specific 
levels of physical conditioning versus the desire to minimize exercise 
i! time and maximize avallaole work time. The operational Issue Is to 
;«f' define exactly how much the exercise time can be reduced with the par- 
^ . ticular protocol without significantly Increasing the risk of Irrevers- 
. Ible physical deterioration. A design factor might be to design and 
V. validate a different exercise technique such as an on-hoard human cen- 
trifuge, which would offer the potential of being a more efficient ex- 
ercise procedure than that used In the original protocol. 

* If such a centrlfugatlon procedure were determined to be ef fect- 

' Ive as an aid to Improving crew productivity, then further trade stud- 

ies wou'id be required to establish the cost effectiveness and the Im- 
pact on mission success of the new Implementation technique. 

; _ Another example might be the effect of whole-body showers on 

crew productivity. Factors that should be considered in a study of 
. V this issue might include: 

- i - The proliferation of skin-surface microflora in the absence 

- * of whole-body showers ~ when using hand and face wash and 
■■ * wipes or cloths only. 

• The relative amounts of skin Infections with and without 
whole-body showers. 

' • The effect of whole-body showers on crew morale. 

I * The amount ot water used per unit of shower time and the 

number of showers allowed per mission. 

• Development, production, and installation costs of wholebody 

• The impact of whole-body showers on Space Station water 
) logistics. 

Still another example might be the effect on the number and 
location of commodes on crew productivity. Examples of factors that 
should be considered in the study of this issue Include: 

• The number of commodes relative to the number of crew mem- 
bers on each work shift. 

• Procedures for commode use in case of failure of one or more 

• Emergency provisions for urination or defecation when com- 
- , mode is unavailable. 



• *^^j£i:' 



Whether or not commodes should be located In other pres- 
surized modules as well as In the habitation module(s) 
(e.g., In laboratories, such as an animal research lab In 
which the crew may wear specialized clothing restricted to 
the lab area) . 

The Intent at this point Is to ask the questions: (1) Will 
changes In the crew operational activity or In the crew organization 
Influence performance and thus Impact productivity?; and (2) Will 
changes In the design concept Influence crew performance and thus Im- 
pact productivity? If productivity adjustments could result from posi- 
tive answers to questions (1) or (2), these become Issues for further 



"fhe criteria of performance , c ost . and mission success proba - 
bility (program confidence) are the principal factors that program or 
project managers and system engineers use In selecting the optimum ap- 
proach to meeting mission objectives. The decision maker must base his 
Judgment on knowledge that a particular Implementation option can or 
cannot achieve the desired productivity levels and meet the basic per - 
fo rmance requirements. In many cases, more than one Implementation 

option can meet the performance requirements, and It Is then necessary i 

to examine the relative costs and success probability associated with 
each approach. ', 

Pe rformance Criteria i 

With regard to performance , the limiting factors on direct human 
involvement are primarily associated with sensing (whether stimuli are 
w1th1n or outside the range of human sensor capability); Information 
processing (whether or not the complexity of the Information to be pro- 
cessed requires supplemental aids); and action (whether or not the ac- 
tion required Is within the range of human motor responses and time ' 4 
availability). ^ 

In a current study being performed by MDAC for the Marshall 
Space night Center entitled, "The Huiran Role 1n Space (THURIS)'"[li, 
37 generic classes of activities have been defined that, when combined 
In the requi d operational sequences, can be used to describe a broad 
spectrum of potential space programs. For each of these activities, 
the limiting factors In terms of sensing, Information processing, and 
motor action--, have been defined, and the requirements for human In- 
volvement nave been described. This generic set of activities was de- 
fined by examining a number of space programs and projects. Including 
Skylab, Space Platform, Space Station Missions, and a Life Sciences 






Each uf these activities could conceivably be accomplished with 
different levels of human Involvement ranging from direct manual opera- 
tion through various degree of automated or robotic support, In the 
THURIS study, five reference steps were Identified along the continuum 
ranging from direct manual operation at one extreme to completely auto- 
mated operations at the other. These steps were: 

Manual - unaided IVA/EVA, with simple tools. 

Augmented - amplification of human sensory-motor capabili- 
ties (powered tools, ej;o-5keletons, etc.). 

Tele operated - use of remotely controlled sensors and actua- 
tors allowing human presence to be removed from the work 
site (remote manipulator systems, teleopera'.ors, telefac- 

Su pervised - replacement of direct manual control of syscem 
operation with computer-directed functions, although main- 
taining humans In supervisory control. 

Independent - basically Independent self-actuating, self- 
healing operations, but requiring human Intervention • cca- 
slonally (automation and artificial Intelligence). 

As a general statement, response time was found to be the most 
generally applicable discriminator between the manually controlled 
modes and the more automated supervised and Independent modes of opera- 
tion. If responses In time periods of seconds or less are required, 
then the activity Is generally best performed In the supervised or In- 
dependent modes. Applications where speed of response would dictate 
that the activities be performed In the supervised or Independent modes 
might Include launch abort procedures and orbital trajectory correc- 
tions. If allowable response times become minutes or hours, and If the 
spacecrew member Is not already loaded then all modes might be appli- 
cable and the criteria of cost effectiveness or <^!rcpss probability 
would provide the more appropriate basis for selec ' <" u particular 
mode of Implementation. 

Co st Criteria 

In considering the Issue of cost . Important fc:^ t^e num- 
ber of times a specific activity Is to be performed, . iv^iber of 
different activities that are required to be performf the .ooe'a- 
t'lOnal sequence. 

Conventional wisdom would suggest that even If a given activity 
were capable of being performed In a manual mode, the cost of a manhour 
or man-minute In space Is so high that If the activity were required 
to be repeated a number of times, a crossover point would quickly be 
reached where It would be most cost effective to Implement a more auto- 
mated approach to the activity performance. In similar fashion. It can 
be reasoned that the human operator Is basically a single-channel 


' ■ 'r-.-^ 


r':jv s 



mechanism and cannot be expected io perform multiple activities simul- 
taneously, although the activities might be performed serially If the 
per*^ormance time permits. 

MDAC has developed costing models .nat provide comparative data 
on the relative costs for each man-machine mode In performing single 
and multiple activities, from one to many thousands of times as a func- 
tion of the time required to perform the specific activity. Figure 1 
summarizes these relationships for three time Intervals In terms 2f 
normalized "Accounting Units"* 

From Figures lA, IB, and IC, a rather significant observation Is 
that the cost level for direct human Involvement (manual, augmented, or 
teleoperated modes) gene'"i''v remains consldeiably lower than the cost 
for remote human Invo' vt.«ent v oervlsed and Independent modes) over a 
large number of times that the a'..lv1ty might be performed (1 to 10,000 
timec). As may be notM, the coit differential's span two orders of 
magnitude when only a few actlvatlors are required (1 to 10) but narrow 
to one order of magnitude when the number of activations approaches 
1000. For most activities, the manual modes are relatively 
Inexpensive. E'^en a space stati , facility charge of $3<?,500-per man 
hour, although a slqnlfUani t.v ,or If lengthy times are Involved, Is 
still a relatively small cost factor until the frequency of use 
approaches 1000. Performing activities In the Independent, super- 
vised, or teleoperated modes requires. In most cases, a 'elaMvely ex- 
pensive Initial Investment In support equipment and software, which 
does not compare favorably with the manual mode unless arrortlzed over a 
large number of uses. 

The -"^re activities t^at are required to accomplish a specific 
mission objt-ctlve, the more time required and the higher the cost. 
This Is true of the manual, augmented, and teleoperated modes of 
operation. In the case of the operational mooes, where human 
Involvement Is more Indirect (I.e., the supervised ground, the super- 
vlsed-on-orblt, and the Independent modes), ttie principal contributor 
to the cost of performing a set of activities 1s more directly depend- 
ent on the cost of the resources and the supporting equipment Items 
required to perform each activity In orbit than on the time required to 
accomplish the activity. This means that In the modes requiring Indi- 
rect human Involvement, the cost reduction due i the potential of 
sharing common equipment Items and common resources can be a signifi- 
cant factor In the cost equation. Figures 10, IE and IF Illustrate the 
cost relationships wheri multiple activities and multiple repetitions of 
a single activity are considered. 

In the- examples shown In Figures ID, IE and IF If only '.'re 
activity were required to be performed. It wo;'ld need to be repeated 
thousa ds of times before "t would be cost effective to provide some 
degree of automated support . On the other hard, If a total of 10 

*An "Accounting Unit" Is the cost to p -form an activity one 
time In the Manual Mode. 




Figura 1. Comparative Costs of Alternitiva Man-Machino Modei for Thraa Diffarent Performance Timet 


1 M'n'jte 





1 10 100 1000 10000 

Number ct T.ines Ai iivitv s Performed 


I 35 

i 30 


t 25 


t 20 

Q 15 


^ 10 

E 5 


/Manual x^ 

1 lO 100 1000 10000 

^'umber of Repetitions of Each Act' 


10 Minutes 




? 100 

Supervised Grouncj. 

Independert v 
Supervised - On Orbity \__ . 


1 "0 100 1000 1C0OO 

Nurr >er of Times Activity is Performed 


- 3(. 


t 20 

Q I"; 

< Teieopera.^d 


1 \ 

Augmented \ \ ^ 

V« --'Supervised 

i ^'l \ \ ^-. 

P 5( -f^anuai "•,,_ *,^ 

1 iO 100 100C 10000 

Number of Repetitons of Each Activity 



100 M.nutes 




Indepi ident 
r perviied Ground 
Supervised - On C bit 

Teleoperated - ^: 

..♦ — • .i' 



1 Ur" I- 

1 10 

L>J 1 I I I I til 

100 1000 10000 

Number of Times Activity is Performed 

Relative Costs 
for Single Actiyil» 


? 35 

z 304- 

c ---' 


£ 20 

? 15 


t 10 

^ 5 


y M ani--' 



1 10 100 1000 10000 

Number of Repetitujns of Eacb Activity 

Regions Wnere Modes are Most Cost Effectn 
When Mu'iipie Activities Required 




*^'e>^^^^ ^^.: . ""' ^ U 

activities have tc be performed to accomplish the mission objective, 
and 1f one of the 10 takes 100 minutes and has to be performed TOGO 
times, designing the mission objective to be accomplished In the 
Independent mode becomes an attractive option. 

Mission Success Criteria 

In developing an estimate of success probability , two Issues 
must be considered. One Is the Issue of human reliability and how the 
human can best be used to reduce risk; the second Issue Is the Impact 
of the state of technological readiness on mission success. Techniques 
are available for estimating the level of technological readiness; the 
use of the human to enhance success probability Is more difficult to 

•'.though precise analytic techniques exist when predicting the 
reliability of complex mechanical or electrical systems with components 
of known reliabilities, and some success has been achieved In predict- 
ing human reliability factors 1n certain well-structured tasks, consld- 
eraole caution must be exercised In attempting to treat analysis and 
Integration of human and machine error In an analogous manner to the 
techniques used 1n dealing with physical systems. The basic problem Is 
that human errors are fundamentally different from machine errors. 

When a physical component falls, the system Is usually designed 
so that the failure Is Isolated and doesn't affect other components. 
When humans make a mistake, resulting frustrations may Increase the 
likelihood of subsequent errors. Machine failures generally require 
human Intervention to repair or replace the failed component. On the 
other hand, humans can monitor their own performance and can often cor- 
rect their own errors before they affect system performance. In physi- 
cal systems, redundant components are assumed or designed to be Inde- 
pendent and, by being placed In parallel networks, can Increase system 
reliability. Redundancy In crew size or presence, however does not 
necessarily Increase reliability and, In fact, the social Interactions 
among crewmembers can lead to common conclusions that may In fact be 
wrong. On the other hand, the human's perception of the likelihood of 
specific components falling can lead to a greater sensitivity and 
awareness for Impending failure and the potential for anticipating cor- 
rective actions. 

While mathematical -modeling of human performance may be possible 
In well-structured tasks, the precise mathematical modeling of human 
performance for systems In the very early conceptual design phase Is an 
elusive goal . 

On the basis of past experience the basic rule when designing 
new systems should be tc consider the human element not 1n terms of 
being a component In series with other system elements and having a 
specific numeric value of reliability, but rather as an element func- 
tioning In parallel with the machine components. The human element can 
enhance system operations by reducing the risk of system failure 
through the use of human performance capabilities to provlot- parallel 

348 r' 



lu i « <!«»' 


'^ or redundant resources In the form of maintenance and servicing, 

repair and replacement, and reprogramming of the machine elements. 


If this philosophy of machine Interface Is adopted, then a man- 

j: ual backup will be provided wherever possible, and the more pressing 

Issue In dealing with mission success probability becomes the Impact 

I that the state of technological readiness of the hardware required may 

<*' have on mission success. 


'^ Lessons learned from the US and Soviet* space programs to date 

suggest that (1) systems can have Indefinite operational lifetimes In 
space If they are designed to permit the contingency of In-flight 
repair and maintenance; (2) structures too large to be launched Intact 
can be constructed and assembled on orbit, using man's unique capablll- 
■ ties; and (3) the flexibility and creative Insights provided by the 

»% crew In situ significantly enhance the probability of successfully 

;1 achieving mission objectives. 

c The ability of the crew to manually assemble delicate Instru- 

■f ments and components and to remove protective devices such as covers, 

T lens caps, etc., means that less rugged Instruments can be used as com- 

-- pared to those formerly required to survive the high launch accelera- 

i- tlon loads of unmanned launch vehicles. As a result, complex mechan- 

isms secondary to the main purpose of the Instrument will no longer 
need to be Installed to remove peripheral protective devices or to 
activate and calibrate Instruments remotely. With the crewmembers 
available to load film, for example, complex film transport systems are 
not needed, and malfunctions such as film jams can be easily corrected 
manually. The time required to calibrate and align Instruments direct- 
ly can be as little as l/40th of that required to do the same Job by 
telemetry from a remote location. 

Specific experiments and operations no longer will need to be 
rigidly planned In advance, but can change as requirements dictate. 
^ One of the greatest contributions of crews In scientific space missions 

can be In reducing the quantity of data to be transmitted to Earth. 
One second of data gathered on SEA SAT, for example, required 1 hour of 
ground-based computer time for processing before It could be used or 
examined, or a value assessment made. Scientist-astronauts In situ 
could determine In real-time whether cloud cover or other factors are 
within acceptable ranges before recording and transmitting data. 

Astronauts can abstract data from various sources and can com- 
bine multiple sensory Inputs (e.g., visual, auditory, tactile) to 

f *The Soviets have been reported to rely heavily on manned 1n- 

3" volvement In order to repair equipment and subsystems with serious 

t shortcomings In reliable and trouble-free service life. 


^ - ---'^S^ 


»^.'5>,SW vt 

V :! -. 


Interpret, understand, and take appropriate action, when required. In 
some cases the human perceptual abilities permit signals below noise 
levels to be detected. Humans can react selectively to a large number 
of possible variables and can respond to dynamically changing situa- 
tions. They can operate In the absence of complete Information. They 
can perform a broad spectrum of manual movement patterns, from gross 
positioning actions to highly refined adjustments. In this sense, they 
are variable-gain servo systems. 

Thus, with the advent of permanently manned space studies, there 
are alternatives to the expensive deployment of remotely manned sys- 
tems, with their operational complexity and high cost of system fail- 
ure. Long-term repetitive functions, routine computations or opera- 
tions, and large-scale data processing functions can be expected to be 
performed by computers capable of being checked and serviced by crews 
in orbit. Just as they are now serviced In ground Installations. In 
addition, the normal functions of the terrestrial shop, laboratory, and 
production staff will find corollary activities In the work done by the 
crews manning the space platforms of the coming generation. 

The criteria of performance , cost , and mission success probabil - 
ity (program confidence) are the principal factors that program or pro- 
ject managers and system engineers use In selecting the most cost- 
effective approach to meeting mission objectives. While the final 
selection In the productivity tradeoff between performance, an 
acceptable probability of success, and the resultant cost must rest 
with the decision maker, these criteria provide a frame of reference 
from which rational or Informed decisions regarding the Impact of 
productivity changes on total system effectiveness can be made. 


[1] McDonnell Douglas Astronautics Company, The Human Role 1n Space 
- Final Report (3 volumes), Marshall Space Flight Center, (Con- 
tract No. NAS-8-35611) October 1984. 

Harry L. Wolbers, Ph.D. Is currently Manager - Manned Systems 
Design, Space Station Programs at the McDonnell Douglas Astronautics 
Company. He has participated In every MDAC Space Station study since 
1963. He Is also a Professor (Adjunct) In the Department of Indurtrlal 
and Systems Engineering at the University of Southern California. 







■ 'if 


HUMAN PRODUcnvrry in space station 




J-^3 i - S / 


Claudia Bird Schoonhoven 
Department of Organization and Management 
San Jose State University 
San Jose, California 95192 



This paper offers a research agenda for analyzing organizational problems 
in permanent organizations in outer space. The environment of space 
provides substantial opportunities for -organizational research -> as we 
face questions about how to organize professional workers in a 
technologically complex setting with novel dangers and uncertainties 
present in the Immediate environment. Although organizational theory and 
behavior have always had important implications for research on human 
habitation in outer space, research into these issues has been limited 
because technological and medical issues have been viewed as more critical 
and thus of higher research priority. It is suggested that knowledge from 
organization theory/behavior has been an underutilized resource in the 
U.S. space program. A U.S. space station will be operable by the 
mld-1990's. Organizational issues will take on increasing importance, 
because a space station requires the long term organization of human and 
robotic work in the isolated and confined environment of outer space, 
vnien an organizational analysis of the space station is undertaken, there 
are research implications at: multiple levels of analysis: for the 
individual, small group, organizational, and environmental levels of 
analysis. The paper reviews the research relevant to organization theory 
and behavior, and offers sug^'estlons for future research. 


The United States is rapidly approaching the beginning of long t<>rm 
habltatior in outer space. A space station will be operational by about 
1993 and space colonization will be close behind. The construction of the 
space station will take place in real time, with the astronauts actually 
assenblying its pre-fabricated modules in space. This achievement will 
mark a very important event in human history, because it heralds the i:nd 
of short term space missions and the beginning of humans living and 
working in space for long periods of time. The ultimate success of this 
remarkable undertaking will hinge considerably on many organizational 
factors. The question of whether human beings can physically survive the 
rigors and real environmental dangers of space appears to have been 

answered favorably with the success of the U.S. and Soviet Union's 




kM iteih. .%'.ii. 


"manned" space flights to date. Whether humans as organizational and 
social systems will survive the transition from work and life on earth to 
long term work and life in a space environment remains a question for 
future observation and research. Years of technological, physical, and 
ne^ al data have been gathered, and yet many basic organlsaclonal and 
managerial questions remain unanswered. 

Why have basic organizational and managerial questions not been 
answered? To a large extent this is because they have not been asked, 
and there are a number of reasons for this, which cross-cut most of the 
social and behavioral sciences including psychiatry and psychology (Santy, 
1983; Helmreich, 1983). (1) A hierarchy of "practical" concerns directed 
the early research. In the early phases of the U.S. space program in the 
1950's and during the post-Sputnik Era, organizational questions were not 
addressed because the question was survival, not how to organize. (2) 
The earliest missions had crews of one and then oerely a few, and mission 
durations were counted in hours. Both of these limitations truncated 
organizational and sociological problems, and thus the apparent need for 
inquiry. (3) An early suspicion of "doctors" ^r>d human scientists 
developed among the astronauts, largely due to the kind of psychological 
and medical testing which surrounded astronaut selection procedures. The 
scenes described in Tom Wolfe's THE RIGHT STUFF and the film by the same 
name provide the reader with colorfv^l i-iages of these early activities. 
Psychiatrists and doctors in general are traditional culprits In the lore 
of pilots, often responsible for grounding military pilots, the group from 
which the early astronauts were recruited. (4) The study of 
organizational Issues surrounding space missions ultimately involves the 
study of NASA as an organization and the astronauts themselves. The 
astronauts are heavily In demand for both training and public appearances, 
and NASA's administrators have been reluctant to increase the demands on 
their time and subsequent good will for the purposes of behavioral or 
social science research. Behavioral research is little understood in 
general, and this has not hastened the case for organizational research. 
(5) Social and behavioral researchers strive to maintain the anonymity 
and privacy of their subject Individuals and organizations. While NASA is 
a large organization, It is nonetheless a single organization. Concerns 
that disguising its Identity would be difficult are understandable. NASA 
is dependent upon government-controlled funds, and thus is appropriately 
sensitive to maintaining a favorable public image and legislative 

All of these issues have Inhibited the development o^ organizational 
and managerial knowledge regarding human behavior in the space 
environment. Nonetheless there is growing recognition of the pressing 
need for this research. Brady, for example, acknowledged that early roace 
missions provided the presumed high motivation among astronauts to 
"..succeed in pioneering a new frontier" (1980). It is an open question 
as to what will motivate productive work bahavior at the high level NASA 
has come to expect, given long term missions in a permanent space station, 
operating under relatively routine work conditions. Sadin has remarked 
that " now seems clear that the stage is set for extending a research 
analysis of interrelated selection, training, and organizational 
problems.." (1982: V-J37). 


yF ■■■\'*s^ 

.•c;jji«j*'.*'4».. "?; 

It l6 Interesting that despite the concerns like those expressed 
above, Bluth has remarked that the U.S. space program has never made a 
formal attempt to establish experiments and record behavioral reactions, 
nor has It used behavioral scientists in its training prograns (1981). In 
contrast, several American scholars have remarked on the extent to which 
the Soviets have undertaken social and behavioral research on their 
cosmonauts. Research findings have been utilized by the Russians In their 
space program to a far greater extent than in the United States (Bluth, 
1981; Connor, 1983). In the sections which follow, we will outline a 
research agenda developed from an analysis of the existing research and 
theory on organization and management as it applies to organizing the 
space station. We will limit out attention to the space station, because 
space colonization involves more global societal organization and 
evolution, and these issues are beyond our current scope. 

Space Sta tion Analogs : Other Isolated, Confined Environments (I.C.E.'s) 

Space stations feasible during the next several decades will be 
fairly small, have limited habitable spaces, and accomodate six to twelve 
human residents for intervals of about three months. The space shuttle 
will service it, transporting limited numbers of people and goods st 
infrequent Intervals spaced to provide adequate provisions. These 
features of a low earth orbit space station in combination with expensive 
co'imunlcatlon linkages to the outside world, combine to produce an 
Isolated and confined environment, referred to as an "I.- C- E." With 
one major exception to be discussed below, relatively few organizational 
and sociological problems among space crew members have been made public. 
Nonetheless, a conceptual analysis of the variables operating within 
future, periLanent space organizations suggests that more rather than fewer 
organizational and structural Issues are likely to be encountered. 
Because a space station will exist within a high danger, hostile external 
environment, with organizational members confined to relatively crowded 
physical spaces with little recourse for escape, space craft will continue 
to be potentially volatile organizational systems. 

Considerable research has been conducted looking at small groups in 
Isolated, confined, and stressful environments (ICE's). These are called 
Earth-based analogs, where analog? are defined as any Earth-based 
simulation or naturally occurlng working or living arrangement which 
replicates in part conditions of space habitation and flight. The 
settings studied have Included Antattic research teams, submarine crews, 
oceanographlc research vessels, Alaskan oil pipeline construction crews, 
and undersea research labs. None of these ICE's exactly replicates 
permanent habitation in space, however, especially the unique stresses of 
space derived from microgravlty and Its related influences. When coirpared 
to the space station, the analogs contain the variable factors of crew 
size, degree of isolation, social and educational background of the crew, 
organization of the crew, nature of the formal work to be done, the 
historical context, degree of confinement, and what are presently 
considered to be the "unique" stresses of space like meteor storms and 
loss of bone density, for example, (Bluth, 1981). They also contain 
variations in organization structure, goals, and expected outcomes. 

Studies conducted on ICE's have Identified a number of psychological, 



-''^Z^^iP '■• ' 

illiiAvA. - 'Wv'.: 

social-psychological and group behavior effects associated with the ICE 
setting, most of which are undesirable from both a mission performance and 
social desirability perspective. For example, on Antartlc stations where 
confinement of mixed scientific and naval crews varies between six months 
to a year, there have been three reported murders (Gloye, 1980). Among 
the navy personel were increases of ttOX In stress-related symptoms of 
anxiety, depression and hostility. Civilian scientists showed the same 
but less intense symptoms. There was a high, consistent emphasis on 
personality-oriented rather than task-oriented behavior (Vlnograd, 197-4). 
Also see G.E. Ruff, 1959; E.K. Gunderson, 1963; C.S. Mullln, 1960; and 
Gunderson and P.D. Nelson, 1963, fo' more detail on the Antaractic 

On oceanographic research vessels differences in educational 
background and formal tasks between ship members appear to be related to 
group disputes and interrupted mission performance. On one, the merchant 
crewmen threw frozen scientific samples overboard, eliminating $50,000. in 
scientific materials, 2 years of scientific Investigation, and one 
doctoral dissertation (Bernard and Klllworth, 1974; also see: Miller, 
Vanderwalker, and Waller, 1971; and Helmrelch, 1971). On Polaris 
submarines on 60-day submerged runs, men reported Insomnia, headaches, 
attacks of anxiety, and depression. Lack of personal space Influenced the 
development of cliques, hostility between them, vulgar language, joking, 
and pecking orders reminiscent of behavior in federal penitentiaries. 
(Sexner, 1968) McNeal and Bluth have summarized the ICE symtomology in 
genuinely hostile environments. These Include: boredom, irritability, 
depression, anxiety, mood fluctuation, fatigue, hostility, social 
withdrawal, vacillating motivation, tension, and sleep disorders. (1981) 

Some would argue that there is high similarity between the ICE 
symtomologies described above and those produced by stress in general. 
The organizational behavior literature on the predictors of stress and on 
the effects of stress on withdrawal, turnover, and other valued 
organizational outcomes is relevant here. Stress has been defined as any 
behavioral responses of an Individual to adversive stimuli (ptressors) 
which push the functioning of the individual beyond ordinary, 
non-emergency coping mechanisms (McNeal and Bluth, 1981). The literature 
on stress management in general may be usefully applied to the space 
environment as a mechanism for deriving potential coping strategies and a 
preparedness for social - psychological problems in space. 

RESEARCH IS NEEDED on the efficaciousness of coping strategies across 
a variety of organizational settings. Including of course, any which are 
concpptually similar to the isolated and confined environment of outer 
space. A major problem with the ICE studies is that they are primarily 
anecdotal, few ^■ariables are quantified, and seldom is there any attempt 
to specify the effects of specific variables on specific other outcome 
variables. Carefully designed research is clearly needed here to move 
beyond the present wholistic conception of an ICE. Future research must 
break out the conceptually independent but potentially interactive effects 
of the variables which combine to produce an ICE. For example, 
psychological withdrawal and subsequent social isolation have been 
reported on the Antarctic stations. Withdrawal is presumed to be a coping 
mechanism for physically isolated settings urder confined conditions. A 


>."*r5!^ ■ 

^i^'^^--i<'iL'^Mki^.. , \^, ; 



-- V 


question to ask Is will an isolated organizational setting where workers 
are not confined continuously produce the same effect? If extensive 
activity outside the confines of the organization were possible, would 
this mediate some of the observed effects? 

There is one additional observation which can be main about the ICE 
studies. Most of the research on long term isolation anu coofinement in 
hostile environments describes behavior and interpersonal relationships 
which do not support optimal mission performance, rtowever. In the largely 
anecdotal ICE research, there Is minimal attention paid to formal 
structure or design of the organization. Typically it is simply ignored, 
often because there is so little variation In formal structure in this 
research with an apparent high reliance on the military command model or 
variations of it in many of the ICE's. While some organizational 
variation has been Introduced by the oceanographic research vessels and 
the north slope settings, the actual structures in place cannot be readily 
ascertained given the reported data. NEEDED RESEARCH: a thorough 
literature review with an analysis of the organizational designs present 
is needed on comparative organizational structures under conditions of 
isolation, confinement, and a hostile external environment as it relates 
to mission performance and the mental and emotional health of the workers. 

The High Funnel: Intra- and Inter-Organlzational Links 

Vfhat also differentiates the space station from other isolated and 
confined organizational settings is what has been referred to as the "high 
funnel" (Hays, 1984). At the present, space missions in the United States 
are closfcly monitored and supported from the ground. While the Intensity 
of ground support is likely to decrease over time with cechnological 
advance, the space station will nonetheless )e part of a very large 
organization with extensive interorganlzational links — metaphorically 
described as a high funnel. In terms of the total organization, it is as 
if the orbiting station itself were at the small end of an upended funnel, 
with the broad end very widely based on Earth. The activities of many 
individuals, groups, and organizations will converge on a very small, and 
for its high cost, fairly modest setting. 

At the narrow neck of the funnel are first mission support and the 
astronaut office, both of which buffe^' and screen inputs to the space 
station, via mission control functions. Expanding from the neck toward 
the base of the funnel is the larger NASA as an organization, with its 
multiple subunits. Spreading widely to the base of the funnel and beyond 
are representatives of a large number of external organizations: 
contractors who built the station, concerned agency and governmental 
groups who fund the space agency, and portions of the scientific and 
commercial communities whose research and commercial enterprises are 
conducted on the space station. This latter mega-group is both quite 
large and not always clearly visible, and yet Its presence will be felt. 
Through communications to and from the space station, technical problems 


•■V*S3P --. 


will need to be solved with tgslstance fvom contractor's engineers and 
decisions will be made and refined regarding research during the mission. 
At this point, one can only specalate as 'o how the high funnel of 
extensive intra- and inter-organiiatlon. relations is likely to Influence 
the work arrangements and work outcomes xn the space station. OOMFARATIVE 
RESEARCH IS NEEDED on the effects of similarly complex "high funnel" 
organizational arrangements on relative organisational power to affc 
declsions, control over valued resources, morale and productivity <• " .hose 
positioned at the top of the funnel itself, the space station equiu^ent. 

There is a fairly deep description of the low power of astronaut 
crews relative to mission control and flight administrators. In former 
Astronaut Cooper's HOUSE IN SPACE (1976), he describes the expectation 
that the astronauts would "..screw something upl", given the constraints 
and technical complexities of the 1973 Sky lab activities. This 
expectation supported the firmly autocratic decision and authority 
structure which governed the astronauts' work. It was also consistent 
with the military command structure the astronauts were familiar with, 
given their mostly ex-military pilot backgrounds. When astronauts were 
scheduled so tightly by the ground that unanticipated events and 
unsyscematically stowed equipment prevented them from maintaining the 
schedule, the three Apollo 3 astronauts closed down communication with 
mission control for twenty-four hours, cleaning and stowing equipment 
properly, awaiting a newly prioritized schedule from mission control. 
This event has become an infamous Harvard Business School case, and titled 
•'Strike in Space" (Balbaky, 1980). This le, of course, one of the only 
publicly known "negative" reactions of astronaut crews during a space 
flight, and it centered on high funnel, organizational issues: the 
relationship between the ground organization and the crew in space. 

If there is an area that is critical to on-station stress. Judging 
from past experience, it is relations with Earth-based personnel. Stress 
from this source in the past has involved a combination of high work 
demands and overspeclf ication of behavior , and these are explicitly 
organizational Issues. However these sources of stress must be dealt with 
and mediated by more appropriate organizational arrangements, because they 
influence quality of life for the space inhabitants. Former astronaut, 
Gerald Carr, reported that "During the Skylab 3 flight, work was mixed 
into the schedule on our days off. Hld-mission we insisted on a full day 
free, -nd that insistence was later labelled rebellion. Like the Russians 
we had our own frustrations with ground support." (Bluth, 1981a) 

When focusing on organization and management issues on the cpace 
station itself, ttie very large Earth-based mocth of the "funnel" could 
easily be overlooked. This would be a mistake, because most of the 
reasons for the space station's existence reside in Earthly motives and 
organizations. Alternatively, one could consider the space station 
mission as a small organization which Includes the immediate ground 
specialists, managers, and communications personnel as well as the crew 
Itself. This space/ground team could receive coHecclve training on 
issues of mutual interest, developing cohesion and trust, and be more 
firmly buffered from outside pressures (Hays and Schoonhoven, 198^). 
Clearly RESEARCH IS NEEDED on the variety of orgaalzational designs and 



Intervention mechanisms potentially applicable -his complex "high 
funnel" organizational setting. 

Organizational Design; Ask the Astronauts 

Research is imperative which solicits the ideas of U.S. astronauts 
and cne earth-based managers of space missions regarding the 
organizational design of future space stations. As of this writing, there 
is no publicly available data In which current American astronauts were 
formally engaged in systematic research as expert iafomants on their 
organizational experiences, needs, and problems while in space. In 
contrast en American scholar has been allowed to interview Russian 
cosmonauts and to publish the results of that research (Bluth, 198.). 

Organizational design is an issue which the Russians have explicitly 
dealt with in their research. In a publicly available, translated 
collection of papers, Novikov reports on Russian organizational structure 
In space: "To our view, the strict distribution of duties and 
responsibilities and refraining from absolute emphasis on a hierarchical 
structure for a crew consisting of 2-3 people and erasure of the concept 
of "commander" is sufficiently expedient as a method which smooths the 
sharpness of such situations. Evidently, one should choose other 
designations which coTespond more to the developed spirit of cooperation 
and fraternity of people who are carrying out important assignments under 
extremely complex conditionc." (1979: 135) 

Stereotypical thinking about the Russian character and apparent heavy 
reliance on bureaucratic organization would suggest that the Russians 
would be among the last societies to reconmiend no commander on a space 
station. Yet, it is their explicit rese trch into the caoject of 
organizational structure in space which led to the rather 
counter-intuitive "no commander" conclusion, above. 

Regarding Oi-ganizational design, there are at least two issues here. 
One is what should the desirable organizational structure be for the crew 
in space? The second is what should the nature of the ground-station 
organizational relationships be. I^la latter question Is usually posed 
within NASA as how much "autonorey" should the crew have from ground 
control. When these questions are addressed, space operations managers 
and adminlstritors are concerned about designing space organizations for 
high levels of performance and for the least cost — the familiar 
organizational concepts of effectiveness and efficiency. While neither 
design question can be simplist Ically answered, many organizational 
theorists would observe that prematurely fixed decisions should not be 
made regarding the proper structure for the space station w^iich are then 
relied upon in an unquestioned manner as "the one besw way" to structure 




and manage a space station. There Is much we do not yet luiow about 
structuring for organizational effectiveness, particularly wtven survival 
rather than merely effectiveness may be the dependent variable as It is In 
the Isolated and hostile external environment of space. Yet, twenty years 
of research on the issue has demonstrated that a single way of structuring 
successful organizations does not exist. Equifinallty emphasizes that 
there is often more than a single path to the same outcome. Several 
satisfactory ways to structure the situation likely exist. These, with 
their differing strengths and weaknesses are reviewed in Sciioonhoven (1983 
and 1984) and crew-based scheduling schemes are (*eveloped in Sims (19BA). 

The systems concept "morphogeny is" reminds us that organizational 
form is capable of adaptation over .me, and yet population ecologists 
argue that organizations are environaentally selected for survival when 
structural inertia exists (Hannan and Freeman, 1984). As yet there Is no 
definitive understanding of the relative strengths of what are presently 
two competing perspectives: adaptive capacity versus structural Inertia 
as they relate to survival of organizations. THIS IS AN AREA OF NEEDED 
RESEARCH. The space station's structure, is presumed to require adjustment 
to changing levels of technology, modifications ir. the number and mix of 
space station missions, shifts in the number and characteristics of crew 
members, and increasing knowledge of space itself. This is a QUESTION FOR 
RESEARCH rather than presumption. 

Initially there will be only one space station, "the" space statioii 
referred to thus far. However several varieties of space stations are 
likely to be developed in the near future, and they vtIII likely specialize 
by mission. There will undoubtedly be purely commercial ventures; the 
enormous costs of space development and missions are expected to be 
shifted to the private sector with industry basically paying :he way in 
the long run. There are also military applications in space and the 
likelihood of stations with predominantly military missions. This 
discussion suggests that different ways of organizi .g are likely to be 
appropriate for space stations as a generic set, depending on which goals 
and missions are pursued, the division of labor, the technology in place, 
and so forth. Whether new ways of organizing will be required which 
deviate from the military command model is an open question, even for 
military space ships. New frontiers often demonstrate the need for 
structural innovations, and research on earth in technically complex 
organizational settings with dramatically different goals is a likely 
target for meaningful research. The recently announced General Motors 
"Saturn" plant is a case in point. 

Were research undertaken in which current astronauts are 
systematically engaged as expert informants, it is important that data 
from participants in the more recent shuttle missions be gathered. These 
missions have been of longer duration, have had larger crews, have had 
greater specialization and division of labor (mission and payload 
specialists, pilots, commander) and have been more technologically and 
relent if ically sophisticated. All of these conditions approach the space 
station conditions more closely than earlier missions have. 

Methods for gathering thifi data should include direct interviews with 
current astronauts who have made shuttle flights £.id with mission managers 



'*■»** J**. .^^ 


for Initial InilghtB Into the major probleraw and varlableb for both 
on-bo«rd and apace aCatlon - tartli organizational laaucK. Tht-ae 
organizational actora are the beat aou-cea of conttaporary and probable 
futvire problema relative to organ'ratlonal laauea. Initial Inttrviewg 
should be eeml atroctured, but open-ended to allow for naxlaua variation 
In responaes. For a conparative baala, astronauts froa earlitr, shorter 
missions with enaller crews will be Important sources of date. A 
representative sample of both sets of crew members can easily be devised 
from the varfooa missions. Getting official access to these laportan: 
actors Is a different Issue, however, Just as organazatlonal entree Is a 
concern Is most organizational research. 

Once this initial davia has been gathered and analyzed for trends, 
explicit hypotheses can be tested using data gathered froE video ici vocal 
tapes of past space missions. Both of these data sources exist. A 
content analysis of such tapes would reveal the extent to which 
perceptions of astronauts and mission managers regarding organizational 
Issues are supported by a larger data base, sampled under varying crew, 
workload, and mission conditions. Video and vocal tapes are a rich data 
source, since they are genuine histories of the actual organizational 

A second, although less ideal approach to hypo'hesls-testlng could 
involve the simulators in which astronauts spend extensive training time. 
The simulators provide controlled environments i*hich reproduce as closely 
as possible selected physical characteristics of the space craft. Given a 
sufficiently realistically simulated work environment, these settings 
could bt jsed to experimentally manipulate organizational characteristics 
in order to study the effects of variation in organizational structure on 
task performance. Flight simulators have been used rather successfully by 
the military and the commercial airlines to train pilots. It appears that 
task conditions are sufficiently realistic that value from the training is 

NASA has funded simulation studies relevant to work in space In the 
past. For example, Brady and Emurian (1978) have studied isolation and 
confinement reactions among volunteer subjects in laboratory settings;. 
However, among the difficulties with simulated space environments and 
other isolation studies is that there is no true danger from the simulated 
space environment to the experimental subjects, and of course a 
microgravlty condition does not exist. Among simulated studies of space 
using long-term isolation (90-days) it has been difficult to obtain the 
experimental conditions desired as a consequence. Crews have high 
awareness of the cimulation, produce a high level of performance, 
generally good morale, and little hostility. The lack of true danger in 
the benign simulations appears to account for generally favorable outcomes 
of the experience. (Seeman and McFarland, 1972) Studies of other 
phenomenon may not depend on the very difficult problem of experimentally 
recreating genuine Isolation and the awareness of confinement, and thus 
:: may be more successf«.'l than the isolation studies appear to have been. 

For example, the simulation of a scientific shuttle missions appears to 
V have been successfully executed, and as a consequence, the organizational 
^ and psychological aspects of mission management were ol)servable. Results 
', suggested that the social dynamics of planning and integrating the 



1 360 

~.Jr ■ '^-'i^!r-^ -.• '■ 



i^fe^K'k^- ■„'■•:,;, ■ ' ' (i) 

components of missions are critical '^o success (Helmrelch, Wilhelm, 
Tanner, Sleber, and Burgenbauch, 1979). 

Naturally, what one simulates depends on the questions addressed in 
the research. Important work is being done in three corners of the 
country in which interactions among commercial airline crew 
lusmbers are analyzed to understand how group dynamics and fonaally 
dif i^rentlated status Influence crew error rates and mission performance. 
Some of these are simulations; others involve direct observation of 
working crews in airplane cockpits, at least initially. This is the work 
of R. Hackman, in process, H. C. Foushee on dyads and triads at 35,000 
feet (1984), and a social psychologist, R. L. Helmreich, long active in 
space-relevant research (1971; 1980; and 1979). What is important is that 
these researchers have creatively examined empirical settings available 
for space-relevant research ^ and in so doing have also made original 
contributions to their bi>se disciplines as well. Collectively their work 
provides a good model for the exciting research which awaits the 
stimulated organization and management investigator. 

Spaci Station as the Ultimate Company Town 

Research is also needed on the interactions between the formal (work) 
organizatior's structure and the social living structure under conditions 
of isolation, confinement, and a hostile external environment. The 
concern is with the social relationships among people who also live where 
they work. The space station has been described as the ultimate company 
town (Schoonhoven, IPS'*: 23). What are the likely interactions between 
off-duty social structure ana the formal organization of the work? How 
does the interacting set of variables influence: (a) mission performance, 
(b) the mental and enotional health of the space workers, and (c) 
individual and colleccive human productivity? Even the Russians have now 
demonstrated their interest in the "emotional enthusiasm" of their 
cosmonauts, because it is presumed to influence "working capacity" 
(Lomov, 1979: 9). 

Recent court cases on Earth involving naval personnel at sea suggest 
that close personal relationships are likely to develop during long 
missions, despite formal organizational restrictions to the contrary 
regarding "fraternization". Privacy needs are likely to develop on long 
term irlsslons as well, which develop fmi both psychological as well as 
social needs. Sexual behavior and tensions must be expected and planned 
for. Studies of all-male prison environments describe in grim detail that 
victimization and destructive interpersonal behavior is common when sexual 
needs are not formally recognized, condoned, and architecturally planned 
for in a facility. 

One may also expect that the formation and dissolution of personal 
relationships is likely to create additional tensions which undermine 
productive, cooperative work group behavior. Obviously the prison studies 
of all-male groups demonstrate that it is not the aoded presence of women 
in an organization which suddenly illuminates this issue. It is clear 
that if officials simply ignore this "delicate" issue, in-mission problems 


are clearly predictable. Slnllarly, if traditional cultural mores are 
presumed to dictate "appropriate" behavior, this nay severely hinder the 
crews' adaptive capacity that will be essential during the longer term 
exploration of the universe and beyond. We would agree with the 
psychiatrist, Santy's conclusion, that "We oust avoid letting rigid 
cultural values and customs, especially In the area of sexual sores, 
prevent research in this area." (1983: 522) 

The existence of truncated social roles is a variable which may 
influence mission performance. Astronauts are currently seen 
unidimensionally as workers and while In space do not enjoy the usual 
variety of social interactions as sybllng, spouse, child, parent, or 
jogging partner. In the formal work group one Is constantly evaluated. 
In a family or friendship setting. Individuals are accepted more 
completely and performance pressures are reduced substantially. Bluth 
remarks that in small confined groups all aspects of dally life are in the 
presence of the same group of people who are always there. The group must 
function for both work and friendship, evaluation and nonevaluatlve 
support •(1981b) Destablized self-concepts can result from Imposed role 
restrictions, which can in turn Influence mission performance. While this 
is an issue for research, prior work In total institutions like prisons 
and state mental institutions suggests that extensive visitation rights 
from family members play a part in depressing undesirable social behavior. 
Similarly having the equivalent of an open-phone to family and friends, 
like E.T.'s phone home, is likely to reduce confinement perceptions by 
introducing more varied role opportunities, alternative sources of 
evaluation, and significant others for self-concept stabilization. This 
is speculation, however, and systematic evaluation of existing mediated 
communications research would be useful here. 

Crew Selection: Structure versus Personality 

Most of the research sponsored by NASA on crew selec* in has been 
conducted from a psychiatric or psychological perspective, because an 
astronaut's psychological stability was considered of primary importance 
during the selection process. (See Santy, 1983, and Helmreich, 1983, for 
reviews of this literature from the two perspectives.) Because the 
disciplines of psychiatry and psychology both emphasize the individual, 
most of the research relevant to crew selection in these two disciplines 
recommends selecting for desirable personality attributes, where 
androgonous, flexible people-oriented individuals are argued to promote 
group cooperation and conflict reduction, for example. TV.ore is evidence 
that systematic characteristics of groups, like their sex ratio for 
example, operate structurally to produce extremes of performance and 
reduced group effectiveness. This research suggests that sociological 
characteristics f^f groups, not merely personality characteristics, may 
strongly Influence desired outcomes on space missions. 

Work groups can be described by the extent to which they are 
numerically skewed by the balance of a dominant, majority social group to 
a numerically rare, minority social group. When there is a low proportion 
of the numerically rare, minority social category In the group, a 
condition of tokenism is said to exist. It has been well documented that 



tokenism (whether intended or Inadvertant) results in high psychological 
stress for the token members; that p-arformance extremes of over and under 
achievement are likely; that inbalanced and unstable self-concepts develop 
for the tokens; that the maintenance of satisfactory social relationships 
within the group is problematic; that dominant members of the group suffer 
humiliation when out-performed by the minorities. This latter dynamic 
undermines the group's ability to encourage excellent perforeaace among 
all of its members. The consequences of tokenism are reduced perforisance 
within the group as a whole. The dynamics of tokenism also lielp to 
perpetuate a system which keeps members of the token's category in short 
organizational supply. This is a structural effect which operates within 
the social system. Independent of selected personality characteristics. 
The decision to send crews into space "token" members in their crews 
can have the negative consequences described above. Sally Ride, Che first 
U.S. woman astronaut in space, and Commander Bluford, the first U.S. black 
astrosdut, both flew under structured token situations. Since the space 
station will present conditions of long term isolation and confinement, 
the presence of numerically skewed crews could have serious negative 
consquences for fut.ire space missions. 

Based on Kanter's work (1977a), research should investigate when a 
work group moves from a numerically skewed, rare token-dominant majority 
structure to a balanced position. This is necessary to inform managerial 
decisions regarding the appropriate gender, ethnic, and cultural balance 
of astronauts necessary to crew successful space station missions. This 
wilJ facilitate recruiting, selection and development decisions for the 
astronaut corps, and the knowledge will help NASA mission? avoid the 
negative consequences of the dynamics of tokenism: psychological stress; 
the performance extremes of over and under achievement; more balanced 
self-concepts; enhanced interpersonal interaction between men and women 
without the pressure of public humiliation of the dominants and consequent 
negative affect; the swings of either social isolation or public scrutiny, 
high visibility, and lack of privacy experienced by tokens. 

It is not enough to recruit for androgenous, psychologically stable 
individuals as has been suggested in some research on the effects of 
isolation and conflnsoent in hostile environments. Some social 
categories, like gender and race, are still very important in our society. 
Women, or members of any other underrepresented category, need to be added 
to total group membership in sufficient proportions to counteract the 
effects of tokenism. What the precise numbers are is a matter to be 
researched. Similarly, if none of the negative effects of tokenism have 
emerged during shuttle flights, then that is also Important information. 
If the negative effects of tokenism have not emerged, perhaps 
inadvertantly avoided through NASA's extensive training progra 
ms , then the specific variables responsible for the depression of this 
structural effect should be understood. It is not enough to be fortunate. 
Expei sive long term missions require maximum information from which to 
make crew assignments. 




Human - Robotic Interaction; Who's In Control? 

One of the objectives of the Space Station Technology Steering 
Committee has been to establish the desired level of technology to be used 
in the Initial design and operation of an evolutionary, long life space 
station and the longer term technology for application to improved 
capabilities (R. Carlisle, 1982). Research is currently in proces to 
develop robotic systems to enhance human productivity in the space station 
context as one avenue for realising this objective. Since robots and 
robotic systems are presently relatively youthful and developing 
technologies, few organizations have had extensive time-based experience 
with them. Little serious research has been conducted to document the 
effects of robots on human performance, safety, productivity, and overall 
organizational performance. 

The installed base of robotic, flexible manufacturing systems has 
grown substantially in recent years in the U.S. and in Japan. These 
Industrial settings provide naturally occuring variation in robotic 
systems and organizational arrangements. They could be used to study a 
number of issues Important to the space, station's development. Of first 
concern is the conditions which promote tafety of the human participants 
in the system. Other pragmatic concerns Include: what has been the impact 
of these technological innovations on %rorker characteristics, including 
required skill levels and educational attainment; how has the relationship 
between managers and the system operators changed with the introductionn 
of intelligent machines; have relationships between members of the work 
group been modified in %»ys which impact overall group and/or 
organizational performance? What are managerial and worker attitudes 
toward Intelligent machines? Do favorable attitudes evolve over time or 
do they require managerial intervention in the form of specialized 
training to promote rapid acceptance and efficient utilization of the 
machines. When we speak of an evolving technological system like the 
space s*"tion, what changes over time ':an be expected in the tasks or jobs 
perfor_;d by humans in the system? TVie basic question to be researched is 
what are the conditions under which strong performance is made possible 
when intelligent machines become a significant element in the 
organizational system. 


In "aeral it appears that existing research and theory on 
organization and manai>eaent can be usefully synthesized to derive insights 
and p ^dictions regarding appropriate organizational designs and probable 
ort,^iili.ationai and small group behavior one can expect on a permanent, 
\uman inhabited space station. However as this review has demonstrated, 
much of the application of these Ideas leaves us with pr«idlctlons which 
should be tested, and with much research before us. Many of the questions 
raised by the relatively novel environmental setting of space can be 
investigatfrl within earthly organizations. The robotics questions raised 
above ar- an example. It would be unfortunate if the U.S. astronauts' 
experience in space along with their counterpart mission managers were 


^•*:-*S»^ -•■•-' . 




overlooked in the Important research ahead. We can no longer rely 
entirely upon the few Russian studies which have been translated and on 
extrapolations from anecdotal, unsystematic research in the ICE space 
analogs for predictions of likely human behavior and mission performance 
in a permanent space station. There are many areas in which the research 
and the applied experience of organization theory, behavior, and 
development are likely to demonstrate high utility in the future success 
of the space station as an organization. If this article has alerted the 
reader to new and exciting research opportunities, then one of the major 
goals of this article will have been achieved. 


[I] Balbaky, M. , "Strike in Space", Harvard University Case Clearing 
House, 1980. 

[2] Bernard, H.R. and P.D. Klllworth. SCIENTISTS AND CREW: A CASE STUDY IN 
COMMUNICATION AT SEA. Office of Naval Research, Contract 
#N0001A-73-A-0«l 7-0001, AD-76 629. Springfield, VA; National Technical 
Information Jervice. April, 1974. 

[3] Bluth, B.J. "Sociological aspects of permanent manned occupancy of 
space", AIAA STUDENT JOURNAL, Fall, 1981, 11-48. 

[4] Bluth, B.J. "Soviet space stress", SCIENCE 81: 30-35, 1981. 

RESEARCH AGENDA. Research monograph reporting the Conference on Hunan 
Behavior in Space, November, 1980, Williamsburg, Virginia. 

[6] Brady, J.V., "Human behavior in space environments, a research 
agenda", in SPACE HUMAN FACTORS WORKSHOP, M. Montemerlo and D.C. Cron 
(eds.), Vol. 2, V-132. 

[7] Brady, J.V. and H.H. Emurlan, "Behavior analysis of motivational and 
emotional interactions in a programmed environment". In H.E. Howe, Jr., 
(ed) NEBRASKA SYMPOSIUM ON MOTIVATION, Vol. 26. Lincoln: University of 
Nebraska Press, 1973. 

[81 Carlisle, Richard, "Space Station", SPACE HUMAN FACTORS WORKSHOP 
PROCEEDINGS, (Eds. M. Montemerlo and A. C. Cron), Vol.1, III-15, August 
24-26, 1982. 

[9] Connors, M.M., A. A. Harrison, and F. R. AJcins, LIVING ALOFT: HUMAK 

[10] Cooper, H.S.F., A HOUSE IN SPACE. New York: Holt, Rinehart and 
Winston, 1976. 

[II] Foushee, H. Clayton, "Dyads and triads at 35,000 feet: factors 
affecting group process and aircrew performance", AMERICAN PSYCHOLOGIST, 



t-^^£i^^'> :V' - ■" ' ' ' ' Cl 

112] Gloye, Eugene. Interview with E.Gloye by B.J. Bluth, based on 
reports like "Alleged murder on Arctic Ice Island is linked to dispute 
over wine: U.S. Jurisdiction in Case Ruled." NEW YORK TIMES, August 6, 
1970, p. 21. 

[13] Gunderson, E.K., "Emotional symptoms in extremely Isolated groups. 
ARCH GE PSYCHIATRY (:362-3t.8, 1963. 

[14] Gunderson, E.K. and P>D> Nelson, "Adaptation of small groups to 
extreme environments". AEROSPACE MEDICINE, 34: 1111-1115, 1963. 

[15] Hackman, Richard, seminar presentations at Stanford University and 
UCLA, Spring, 1985, on "Team performace in aeronautical and space 
environments" . 

[16] Hannan, M.T. and J. Freeman, "Structural inertia and organizational 

[17] Ha>s, Dan, "The high funnel". Unpublished draft. December, 1983. 

[18] Hays, Dan, and Claudia Bird Schoonhoven, "The high funnel: 
Interpersonal design for a space station", paper presented at the Academy 
of Management Symposium, "Organizing a NASA space station; high 
technology work 160 miles above the earth", August, 1984: Boston. 

[19] Helmreich, i'obert L., "The Tektite Human Behavior Program", in 
Miller, Vanderwalker, and Waller (eds), TEKTITE 2: SCIENTISTS IN THE SEA. 
Washington DC: U.S. Government Printing Office, 1971. 

[20] Helmreich, Robert L., "Applying psychology in outer space: 
unfulfilled promises rtvisited", AMERICAN PSYCHOLOGIST, April, 1983: 


[21] Helmreich, R.L., J. W.'lhelm, T. Tanner, J.E. Sieber, and S. 
Burgenbach. "A critical review of Ames Life Science Participation in 
Spacelab Mission Development Test III: the SMD III Management Study. NASA, 
TM 78494. Ames Research Center, Moffett Field, June, 1978. 

[22] Kanter, Rosabeth Moss, "Skewed sex ratios and responses to token 
women", AMERICAN JOURNAL OF SOCIOLOGY, Vol. 82, No. 5, March, 1977. 

[23] Loraov, B.F., "General characterizations of problems of space 
psychology", in Petrov, Lomov, and S&msonov (eds), PSYCHOLOGICAL PROBLEMS 
OF SPACEFLIGHT. Moscow: Nauka Press, 1979. 

[24] McNeal, S.R. and B.J. Bluth, "Influential factors of negative effects 
in the Isolated and confined environment" . Fifth Princeton/AIAA/SSI 
Conference on Space Manufacturing, May IS-?!, 1981. 

[25] Mullin, C.S., "Some psychological aspecvs of isolated Antarctic 
living", AMERICAN JOURNAL OF PSYCHIATRY 117: J23-325, 1960. 

[26] Novikov, M.A. , "Psychological problems of v raining Cosmonauts", in 
B.N. Petrov, B.F. Lomov, and N.D. Sacsonov (eds), PSYCHOLOGICAL PROBLEMS 



*J1!*4.' C-Vj-,;v 

OF SPACEFLIGHT. Mowcow: Nauka Press, 1979. 

[27] Ruff, G.E., E.Z. Levy, «nd V.H. Thaler, "Studies of Isolation and 
confinement", AEROSPACE MEDICiWE 30: 599-604, 1959. 

[28] Santy, Patricia, M.D., "The Journey In and out: psychiatry and space 
exploration", AMERICAN JOURNAL OF PSYCHIATRY 140: 5, May, 1983: 519-527. 

[29] Schoonhoven, Claudia Bird, "Organizational considerations for the 
space station's development", Appeodi)c to Chapter 4, AUTONOMY AND THE 
HUMAN ELEMENT IN SPACE. Technical Report, NASA-ASEE Suoner Faculty 
Research, Stanford University, 1983. (Forthcoming in AUTONOMY AND THE 
HUMAN ELEMENT IN SPACE, NASA Technical Publication, 1985) 

[30] Sexner, J.L., "An experience In •ubmarine psychiatry". AMERICAN 
JOURNAL OF PSYCHIATRY, 1, July, 1968: 25-30. 

[31] Sims, Henry P., "Designing station organizational systems: the 
special case of crew scheduling". Paper presented at Academy of 
Management Symposium, "Organizing a NASA Space Station", Boston, August, 

[32] Vinograd, Sherman P, Project Director. STUDIES OF SOCIAL GROUP 
December, 1974, pp 135-140. 

[33] Wolf, Tom, THE RIGHT STUFF. New York, Farrar, Strauss and Giroux, 


Dr. Schoonhoven Is an Associate Professor of Organization and Management, 
San Jose State University and has been a Visiting Scholar, Stanford 
Graduate School of Business during 1984 and 1985. Her Ph.D. is from 
Stanford University in organization theory and behavior, and she was a 
NASA-ASEE Summer Faculty Fellow, 1983. Currently in the early stages of 
research on success factors in semiconductor start-up companies, she 
specializes in the management of Innovation and organizational design. 


••\"^£5p ■ 







Scott Danford, School of Architecture and Environmental Design 
J&aes Melndl, School of Management 
Raymond Hunt, School of Management 
State University of New York at Buffalo 


The magic, mystery and romance which will surely surround Its 
first years of operation and the careful selection of motivated, dedica- 
ted (to NASA) crews will doubtless combine to ensure high levels of pro- 
ductivity among the first several crew generations to occupy space sta- 
tion. However, once space ^^tation's operations become routine and the 
magic, mystery and romance give way to perceptions of its being just an- 
other isolated and confined place to go to work, high levels of perfor- 
mance and effectiveness could become problematic. It will be at that 
point that the crews' environmental f*.ontext will become an increasingly 
significant productivity parameter. 

In consideration of this, unprecedented Cfor NASA) and commend- 
able attention is being paid to issues of "crew productivity" during 
current design work on space station. Unfortunately, this "crew produc- 
tivity" is being defined almost exclusively in terms of "human factors" 
engineering and "habitability" design concerns. While such spatial en- 
vironmental conditions are, of course, necessary to support crew perfor- 
mance and productivity, they are by no means sufficient to ensure high 
levels of crew performance and productivity on the post-IOC (Initial Op- 
erational Configuration) space station. What is being relatively ig- 
nored is the role of the organizational environment as a complement to 
the spatial environment for influencing crew performance in such isola- 
ted and confined work settings. 

This paper identifies three possible models of operation for 
post-IOC space station's organizational environment and explains how 
they and space station's spatial environment will combine and interact 
to "occasion" patterns of crew behavior - both desireable and undeslre- 
able. The paper concludes by suggesting a three phase program of re- 
search designed (a) to identify patterns of crew behavior likely to be 
"occasioned" on post-IOC space station for each of the three models of 
operation and (b) to determine "proactive'Vpreventative management stra- 
tegies which could be adopted to maximize the emergence of preferred 
outcomes in crew behavior under each of the several spatial and organi- 
zational environment combinations. 


■•■- '■0if^' 





While space station's initial organizational model of operation 
will likely follow the successful traditions established by over two de- 
cades of U.S. manned space programs, legitimate questiontt can be raised 
about its long-term appropriateness for what is likely to be a more rou- 
tine, more commercially-oriented, less glamorous, more isolated and con- 
fined post-IOC (Initial Operational Configuration) space station [3]. 
Once the magic, mystery and romance of space station's initial years of 
operation fade to leave rather bleak but accurate perceptions of its be- 
ing just another isolated and confined place to go to work, one can ex- 
pect the station's environmental context - both spatial and organiza- 
tional - to become an increasingly important productivity parameter [2]. 

The Spatial Environment 

Fortunately, the spatial component of that environmental context 
is receiving unprecedented (for NASA) and laudable attention as a poten- 
tial productivity parameter during the current design work on space sta- 
tion. For example, beyond the expected "human factors" engineering con- 
cerns, NASA has begun exploring "habitability" design concerns of priva- 
cy, crowding, territoriality, personal space, way-finding, spatial ori- 
entation, ccc, which would have been considered frivolous or at best 
largely irrelevant for any of the previous U.S. manned space programs 
(including, unfortunately, Skylab) [4]. 

While design work on space station's spatial environment Is con- 
tinuing (and is therefore unavailable for public presentation at this 
time), the design concept work of NASA's in-house, 300+ member "Skunk- 
works" team done during the Summer of 1984 is available for illustra- 
tive purposes (see Figure 1) [2]. While these design concepts for 

Illustrative Space Station, Port View 

Nlttl«i> H«<al*. 

M.tiiail'o nxtiii 

Ulft 5cUnc», Litorttor, HO<ultV KinrUlt rr«c«ifll>| l.i»»'«««'I *•*«*• 






space station's spatial environment were not Intended to leflect all 
state-of-the-art "human factors" engineering and ''habl tabillty" design 
concerns, they nevertheless give some feel for the degree of attention 
being focused on these concerns as potential productivity parameters for 
the crews of space station. 

But while such features aa "zero-gravity body posture" appro- 
priate design, artificial vertical referencing systems, individual pri- 
vate crew quarters, etc., are obviously necessary to support long-term 
crew performance and productivity, they are still not sufficient to en- 
sure such behaviors. While the spatial environment's influence on users' 
behaviors is undeniable, there are few who would argue the case in favor 
of some form of spatial environment (I.e., architectural) "determinism." 
Rather, the spatial environment is instead viewed as "occasioning" cer- 
tain patterns of user behavior - but even then only to the degree that 
other environmental factors are at least not working at cross purposes 

Consequently, the design of the environmental context in which 
podt-IOC space station's crews must operate *d.ll have Co address consid- 
erably more than "human factors" engineering and even ''habitablllty" de- 
sign spatial concerns if high levels of crew performance and productivi- 
ty are to be not just "enabled" but actually "occasioned." Providing a 
"productivity occasioning" environmental context for the post-IOC space 
station will necessarily mean considering those other envircimental fac- 
tors as well. And foremost among those relatively ignored "other" e.ivl- 
rorniental factors in terms of its likely Influence on crew performance 
and productivity in isolated and confined settings like the post-IOC 
space station is the organizational environment. 

The Organizational Environment 

Combining and Interacting with post-IOC space station' a spatial 
environment features to either compromise or reinforce its "productivity" potential will be a number of organizational environment 
concerns. Regretably, despite its virtual certainty as a productivity 
parameter on-board space station, the model of operation that organiza- 
tional environment will adopt remains largely undefined and relatively 
ignored. As a significant complement to post-IOC space station's spa- 
tial environment for influencing crew performance and productivity, one 
might have expected more attention to have been paid to this area - in- 
cluding the possibility of having the organizational environment's se- 
lected model of operation serve as one of the prime drivers for the spa- 
tial environment's design to ensure (at minimum) their compatlDility. 

Unfortunately, because the on-going design of space station's 
spatial environment continues largely outside of any major consideration 
of what model of operation its organizational environment should ideally 
adopt, the selected spatial environment will inevitably become the limit- 
iig factor in any eventual choice of organizational environment concerns. 
But while the opportunity to have post-IOC space station's organizational 
environment serve as a primary consideration in the design of Its spatial 
environment has been lost (with the consequence that certain spatial and 
organizational environment combinations with strong "productivity occa- 



lteit*^.t^>»ir>i^'-'~'. V 

—^ l il i < I MH !M' WI'» I W )>Mi 

stoning" potential may no longer be possible), there still remains the 
choice of an organizational environment which can hopefully complement 
the selected spatial environment. 

Assuming that the selected spatial environment will possess the 
necessary "human factors" engineering and "habltablllty" design features 
without which crew functioning would be severely affected (If not under- 
mined altogether) , there still remains certain significant (even if con- 
strained) performance- and productivity-relevant decisions to be made a- 
bout post-IOC space station's organisational environment. Even if the 
selected spatial environment were to present "human factors" and 'habi- 
tability" features permitting potentially high levels of crew per ormance 
on space station, it would be the organizational environment which would 
largely determine the degree to which those potentials were or were not 

Indeed, with space station likely to be a rather spartan spatial 
environment for a variety of budgetary, practical or oversight reasons, 
the burden on the organizational environment to realize fully whatever 
potentials that the spatial environment provides will be enormous if 
post-IOC s[ ace station is to entertain any thoughts of being a productive, 
cost-effective enterprise. Consequently, thf> design of space station's 
organizational environment, although perhaps not as significant as it 
might have been had it served as one of the prime drivers for the design 
of cipace station's spatial environment, nevertheless remains a critical 
influence on space station's crew performance and productivity. 


Because space stat con's organizational environment remains large- 
ly undefined and relatively ignored, questions about which models of op- 
eration might bfc considered to reinforce the chosen spatial environment's 
"productivity occaLioning" potential remain unanswered. While some might 
favor a continuation of the successful operational traditions established 
over the past two decades of U.S. manned space programs, such proposals 
would seem to ring untrue due to the numerous ways in which post-IOC 
space station will be substantially different from any of NASA's previous 
experiences - including the Skylab program. With post-IOC space station 
likely to provide a more routine, more ccmnercially-oriented, less glam- 
orous, and more isolated and confined experience for its crews than will 
initially be the case when space station first becomes operational, the 
appropriateness of continuing to carryover traditional operational pro- 
cedures from NASA's previous manned jpace programs can logically be 

The "Paramilitary" Model of Operation 

The practice of maintaining an organizational environment de- 
signed for NASA's mor j traditional "paramilitary" model of operation 
(i.e., NASA/military astronaut-dominated operations in which any commer- 
cial activity is performed primarily by tht crew on behalf of earth- 
bound clients) may be already becoming obsolete as the private sector 




HmfkMiai. -v^K, 

begins to book o-.ijage for its own employees on jhuttle (STS) missions 
so tiint they can conduct their own proprietary experiments. Once space 
station passes through its fir t few "shake-do-r." years, the incidence 
of private sector employees going to work in space to conduct experi- 
ments and even to engage in manufacturing activities will likely become 
the "norm" rather than the current "exception." At that point, NASA's 
traditional "paramilitary" model of operation could begin to constrain 
such activity and thereby compromise the commercial potential of post- 
IOC space station. Consequently, organizational environments based upon 
alternative models of operation should at least be considered. 

The "Host" Model of Operation 

One moderate alternative to the "paramilitary" model and one tor 
which there are ntimerous earth-bound illustrations is the "host" model. 
Applied to the post-IOC space station, this model would find a smaller 
NASA/military crew functioning 6s "host" to on-board civilians engaged 
in conmercial/proprietary activities on behalf of private sector, earth- 
bound jorportations. In such a model of operatio- ;.he crew would retain 
full authority over all on-board activities analogous to th authority 
vested in the crew of a commercial airliner or passe ^&t ship but would 
generally be reluctant to exercise such duthority except under the most 
extraordinary conditions. Obviously even this moderate alternative to 
the traditional "paramilitary" model would necessitate a substantially 
different organizational environment for post-IOC space station to that 
which NASA has provided to date. 

The "Corporate" Model of Operation 

A second alternative representing a somewhat more radical depar- 
ture from NASA's traditional "paramilitary" model of operation would be 
the "corporate" model. Characterized by NASA's removing Itself com- 
pletely from day-to-day operations to allow the post-IOC space station 
to become a predominantly civilian, commercial enterprise engaged in 
proprietary research and manufacturing, this model or operation would 
see NASA's role return to a focus on advanced research and development 
activities in keeping with its original charter. Under this "corporate" 
model of operation, post-IOC space station could become a government 
leased commercial facility with NASA's involvement restricted to pro- 
viding selected technical services to private sector occupants on a "by 
request," fee-paid basis. With NASA's withdrawal from any routine op- 
erational roles on-board, post-IOC space station's organizational envi- 
ronment would obviously be a radical departure from that to be found 
for a "paraiuilltary" model of operatron. 

Consequences of the Choice of Model 

For each of these three models of operation for post-IOC space 
station's organizational environment there are hot' costs and opportu- 
nities which make the choice between them less intuitively apparent 
than it might at first seem. The only thing that is intuitively ap- 
parent is that each of these three models of operation for the oreani- 
zatlonal environment - when combined with space station's selected 
spatial environment - will inevitably "occasion" markedly different 


-■"r^tSf'" ■ 


patterns of crew behavior and thereby differentially impact crew perform- 
! ance and productivity. 

While the spatial environment will define upper and lower limits 
J. on the crews' performance and productivity potentials by its accommoda- 

tion (or lack of accommodation) of the crews' "human factors" and "habi- 
tabllity" needs, it will be the organizational environment's adopted 
7 model of operation - interacting with the selected spatial environment - 

■^ which influences the degree to which those potentials are to be realized. 

^ By its reinforcing or compromising spatially defined "occasions" for 

I various patterns of behavior which will influence the crews' performance 

J ^nd productivity, the organizational environment will necessarily become 

JF a significant productivity parameter on the post-IOC space station. 



Consequently, the choice between alternative models of operation 
for post-IOC space station's organizational environment cannot be mac"" 
arbitrarily, intuitively or based upon some established tradition. The 
potential consequences for post-IOC space station's crew performance and 
prodi' tivity .ii.e too signliicant to base the choice on anything less rig- 
orcuL or empirical than that which would be expected for comparable "hu- 
man factors" engineering decisions. 

Beyond the Choice of Model 

Moreover, while it is important to realize the potentially signi- 
ficant consequences of choosing between alternative models of operation 
due to the unique opportunity which each - when combined with the -^elec- 
ted spatial environment - presents for influencing 'occasions" for spe- 
cific patterns of crew hehavior, it is equally important to realize that 
each spatial and organizational environment combination actually "occa- 
sions" a range of both desireable and undeslreable patterns of crew be- 

Because there is no reason to expect there to be some special or 
magical combination of spatial and organizational environments for post- 
IOC space station (or any setting for that matter) which will always de- 
liver the "goods" and invariably suppress the "bads" in "occasioned" 
crew behavior, steps will surely have to be taken to influence the dis- 
plays of certain of those behaviors- For that reason, management stra- 
tegies for encouraging desireable patterns of "occasioned" crew behaviors 
and discouraging undeslreable patterns of "occasioned" crew behaviors 
will need to be identified for each of the spatial and organizational en- 
vironment combinations before any Informed choice can be made between al- 
ternative models of operation for post-IOC space station. 


Even though the era of large scale commercial research and manu- 
facturing acrivlty for space station is still wtll over a decade away by 
even the most, optimistic estimates, it is Important to note that it is 
already too late for any particular model of operation for post-IOC space 
station's organizational environment to be included among the considera- 



tions influencing the current design work oi space station's spatial en- 
vironment. Because of that lost opportunity, organizational environment 
options for post-IOC space station are already constrained to those which 
can be accommodated by the selected design for space station's spatial 

Yet while possibilities for certain spatial and organizational en- 
vironment combinations may have already disappeared, there still remain 
important (even if co.istrained) options for post-IOC space station's or- 
ganizational environirent which will still significantly influence crew 
performance and productivity. To enable informed choices among those re- 
maining organizational environment options, a three phase program of re- 
search will be needed not only to understand the patterns of crew behav- 
ior "occasioned" by each spatial and organizational environment combina- 
tion but also to identify effective "proactive"/preventative (as opposed 
to "reactive"/corrective) management strategies for both encouraging and 
discouraging certain of those behaviors so "occasioned." 

Occasioned Crew Behavirrs 

Because the chosen model of operation for post-IOC space station's 
organizational environment - when combined wi»-h the selected spatial en- 
vironment - will significantly influence patti ms of crew behavior, it 
is suggested that Phase One of this suggested program of research first 
identify those patterns of crew behavior "occasioned" by each of the spa- 
tial and organizational environment combinations and then determine the 
desireability or undesireability of all such "occasioned" behaviors in 
terms of their impact on long-term crew performance and productivity. 

This could be done through full-scale, earth-bound behavioral sim- 
ulations involving longitudinal study of task performance and effective- 
ness for each of the alternative spatial and organizational environment 
combinations. From such research one would know the patterns of crew be- 
havior likely to be "occasiuned" by various spatial and organizational 
environment combinations and chose "occasioneu" behaviors' probable im- 
pact on long-term crew performance and productivity. 

Management Strategy Options 

With each spatial and organizational environment combination for 
the post-IOC space station "occasioning" a range of crew behaviors - 
both desireable and undesireable in terms of their impact on long-term 
cvci performance and productivity - it is suggested that Phase Two of 
this suggested program of research identify manajiament strategy options 
which organizations operating in close earth analogs to each of the spa- 
tial and organizational environment combinations have employed in deal- 
ing with both desireable and undesireable patterns of crew behavior. 

By examining readily available, documented experiences in these 
close earth analogs (e.g., extended underwater cruises in nuclear sub- 
marines for the "paramilitary" model; "wintering over" in Antarctica for 
the "host" model; and isolated Alaskan oil pipeline camps for the "com- 
mercial" model), a range of options for influencing patterns of "occa- 
sioned" crew behaviors associated with each of the spatial and organiza- 



'* m 

tional environment combinations can be Identified. From this research 
one would obtain lists of hypotheticaiiy appropriate and inappropriate 
management: strategy ^options for influencing both desireable and undeslre- 
able patterns of crew behavior associated with each of the spatial and 
organizational environment combinations to be considered for post-IOC 
space station. 

Empirically Informed Choice 

Because an informed choice between alternative spatial and organi- 
zational environment combinations for the post-IOC space station is only 
possible when the hypothetical appropriateness or Inappropriateness of 
alternative management strategy options for influencing both desireable 
and undesireable patterns of "occasioned" crew behavior have been empiri- 
cally tested for each model of operation, it is suggested that Phase 
Three of the suggested program of research determine the effectiveness 

H of the various management strategy options. 

', As in Phase One of this suggested program of research, full-scale 

^ earth-bound, behavioral simulations involving longitudinal study of task 

n performance and effectiveness for each of the alternative spatial and or- 

fganizational environment combinations could be employed to provide an em- 
pirical t€;st of the actual appropriateness and inappropriateness if hypo- 
-^ thetically appropriate and inappropriate management strategy options for 

\}' dealing with patterns of both desireable and undesireable crew behavior. 

^ From this research one would have a group of empirically-based indicators 

, . of the probable effectiveness of various management strategy options for 

! influencing patterns of "occasioned" crew behavior likely to be associ- 

ated with each of the alternative spatial and organizational environment 
- combinations for the post-IOC space station. 



Jnderlylng all this discussion of "alternative spatial and organ- 
izational environment combinations for the pcst-IOC space station" and 
"on what basis one might begin to make informed choices between them to 
maximize crew performance and productivity" are three basic arguments: 
(1) that the crews' environmental context will become increasingly signi- 
ficant as a productivity parameter as post-IOC space station becomes 
just another routine. Isolated and confined place to go to work; (2) that 
the spatial component of that environmental context will define upper and 
lower limits for the crews' performance and productivity "potentials" by 
its accommodation (or lack of accommodation) of their "human factors" 
and "habitability" needs; and (3) the organizational component of that 
environmental context will strongly influence the degree to which those 
"potentials" are ever realized through the model of operation adopted 
and the appropriateness and effectiveness of management strategies em- 
ployed within that model of operation to encourage/discourage patterur 

; of deslreable/undeslreable crew behaviors which (in combination with th-. 

^--. selected spatial environment) it "occasions." 

.; ' Of course, it should be recognized that the aforementioned dis- 



■*-'#*«Sif ■;. --■ 

V *^;i>«». vv.' 

cusslon and basic crguments are rooted in the theoretically acceptable 
but practically improbable assumption of "all other conditions being 
held constant." Without doubt one can expect that in the "real world" 
of post-IOC space station all other conditions will not be held constant 
and they thereby will introduce an avalanche of uncontro] led complexity 
into this simple model of environmentally "occasioned" bei-avlor and Its 
Influence on long-term performance and productivity. 

But such is often the case with empirical research - one begins 
with an admittedly simple model and gradually adds complexity as pressing 
circumstances dictate and/or as research-based understanding allows. 
That is the rationale behind this simple discussion of alternative spa- 
tial and environmental environment combinations and how one might begin 
to make informed choices between them for post-IOC space station. What 
has been presented is an admittedly simple model of environmentally "oc- 
casioned" behavior and that behavior's influence on long-term crew per- 
formance and productivity - a model to which complexity can hopefully be 
added as appropriate research -based understanding and knowledge Increase 
to warrant more complex model building. 

And that, in large part, is why it is not too early to begin the 
suggested program of research now. It is not that the decade or so that 
stands between now and the post-IOC space station era is such a short 
period of time in any absolute sense, but rather that, given the current 
simplicity of our models of environmentally "occasioned" behavior, there 
is even now barely enough time for the growth and development which our 
models must undergo if they are to deliver the needed understanding by 
the date required. 


[1] Danford, Scott, "Dynamic Reciprocal Determinism: A Synthetic Trans- 
actional Model of Person-Behavior-Environment Relations," in Amedeo, 
Griffin and Potter (Eds.), EDRA 1983: Proceedings of the Fourteenth 
International Conference of the Environmental Design Research Asso- 
ciation , Environmental Design Research Association, Inc., 1983, pp. 

[2] Danford, Scott, Selected Psycho-Social and Organizational Configura- 
tion Drivers and Space Station Desig n, NASA- Johnson Space Center, 
Medical Research Branch, 1984. 

[3] Johnson, Richard and Frcitas, Robert, Autonomy and th . Human Element 
in Space: Executive Summary , National Aeronautics and Space Admin- 
istration, 1983. 

[4] Le ssons Learned on the Skylab Program , NASA- Johnson Space Center, 


,-. ^ 

-*"*-.i)l«P l?»^^ii«.;.^vrcr-r^-,-v<v ^s' ^ 








^ N86-15190 



Chris Culbert 

Artificial Intelligence Section 

Mission Planning and Analysis Division 

NASA/Johnson Space Center 



'^ Artificial Intelligence (Al) is flourishing outside the bounds of its 

f traditional academic environment. A number of the computer technologies 

\ pioneered in the Al world can make significant contributions to increasing 

STS operations productivity. Application of expert systems, natural language, 
speech lecognition, and other key technologies can significantly reduce 
manpower while raising productivity. Many aspects of STS support lend 
themselves to this type of automation. The Artificial Intelligence Section of 
the Mission Planning and Analysis Division has developed a number of 
functioning prototype systems which demonstrate the potential gains of 
applying At technology. 


The field of Artificial Intelligence had its beginnings over 20 years 
ago. The original goal was to make tne computer "think" like a human; i.e., 
to make computers solve problems in a fashion very similar to the way 
humans do. After an initial flurry of interest, the mainstream of the 
computer industry stayed out of serious research into Al, and the work 
continued primarily in the academic environment at such places as Carnegie- 
Mellon University, Stanford University, and M.I.T. Although the original 
problem has proven to be considerably more difficult than initially 
anticipated, significant strides have been made towards improving the 
capability of both computer hardware and software. These developments 
have application to real-world problems, and in recent years, Al has begun to 
flourish outside the academic environment. Spurred by the success of a few 
key technologies, commercial development is placing more and more of the 
computer advances nioneered by the Al researchers into the mainstream 

Artifida' Intelligence is generally split into a number of subfields, 
including: computer vision, natural language, speech recognition and 
synthesis, robotics, and expert systems. Applications in all of these areas are 
in commercial use. Other areas, such as common sense reasoning, true 
computer learning, self-adapting systems, etc., are still not well developed 



outside the research environment. Although numerous universities are doing 
work in Al, most of it is at the graduate level or above, and researchers with 
extensive Al background currently command top dollar in the job market. On 
the commercial side, much of the current effort has been devoted to 
providing software (expert system shells, natural language systems, computer 
algebra) and hardware (vision, speech synthesis, speech recognition, robotics) 
tools which aid in the development of Al products, thereby, eliminating or 
reducing the need for expensive Al experts. Numerous companies have 
formed in the last three years to capture a part of this growing market. 

The most common computer language for Al work is still LISP, 
although PROLOG, C, and Ada have gained some acceptance. PROLOG is 
widely used in both Europe and Japan. There is no universally accepted LISP 
standard, but Common LISP is rapidly evolving into a de facto industry 
standard. Dialects of LISP are available on many machines, from personal 
computers to mainframes. Computers specially designed to run LISP have 
done much to improve the performance cf the language and are available 
from a growing nur'^er of vendors such as Symbolics, LMI, Xerox, and Texas 
Instruments. Compeiicion and increased sales are steadily lowering the price 
of the hardware. These machines also incorporate many software 
environment advances: windows, pointing devices, high resolution bit- 
mapped graphics, and object oriented programming capability. 


The Space Transportation System (STS) is rapidly becoming a fully 
operational program. Many aspects of this program face the chalfenge of 
reducing manpower and resources while still supporting a growing flight 
rate. Meeting this challenge will require increased reliance on automation 
and higher productivity in all phases of mission support. STS operations, 
particularly in the areas of real-time support of the Mission Control Center 
(MCC), system performance analysis, and mission planning have matured to 
the point where a large amount of expertise has been developed. Due to the 
nature of the organization, much of this expertise has been recorded in the 
form of flight rules, procedures handbooks, etc. 

Many of the available Al technologies are directly applicable to these 
NASA operations. Expert systems can be applied to well understood, routine 
tasks to capture specializeci knowledge and reouce manpower requirements. 
Improved man-machine interfaces using voice synthesis, voice recognition, 
natural language, and advanced graphics can increase the productivity of the 

Many of the MCC tasks mvolve monitoring data for potential 
problems or recording information for analysis. Since these procedures are 
now well understoodf, and generally well documented, they readily lend 
themselves *o implementation in expert systems where the specialized 
knowledge can be captured in a computer program. Although systems that 
adapt and respond to new situations as they are occurring are not available 
as of yet, expert systems have proven quite capable in well defined problem 
areas. Expert systems make excellent monitoring tools since they never get 
bored or tired, are always alert, react faster than humans, and can't retire or 






quit. Expert systems can provide a wide range of aid; from merely warning 
controllers of developing problems, to advising controllers on potential 
responses, to actually correcting problems as they occur. The use of expert 
systems for these kinds of tasks could potentially free large amounts of 

Another potential benefit of expert systems is a reduction in the 
effort required to modify the system. Since knowledge is represented and 
coded in a manner more closely akin to the way humans think, theoretically, 
it is easier to maintain the information. This theory has been difficult to 
prove since most expert systems in commercial use are near the beginning of 
their life cycle. Also, verification of expert systems is still a poorly understood 
concept. Current techniques involve "training" the expert system in a 
manner very similar to the way human experts are trained, through 
repetition and simulation of problem situations. For most NASA applications, 
it is anticipated that the expert systems would work in parallel with human 
experts until confidence is gained that the expert system performs correctly 
and at an expert level. 

The increased use of automation in all functions will place a larger 
emphasis on the manner in which humans interact with the computer. The 
man-machine interface must become more flexible and less dependent upon 
the users learning specific procedures and peculiar syntaxes. The emphasis 
will be on allowing users to work with the computer tools in a manner that is 
comfortable for the user and not dependent on hardware/software 
limitations. Natural language processing could allow users to enter 
commands or information into the system without learning a special 
terminology. Speech recognition systems could allow users to enter 
information while leaving their hands free and also avoids typing mistakes. 
The combination of natural language arid speech recognition provides a 
flexible yet powerful manner for users to enter information into the 
computer. At the same time, speech synthesis systems can provide for output 
from the system without requiring direct user attention. Better use of color 
and advanced graphics systems will also improve the effectiveness of 
information presented to the user. 


The Artificial Intelligence Section, Technology Development and 
Applications Branch of the Mission Planning and Analysis Division (MPAD) has 
been actively working on Al applications for over a year. Projects are 
currently under way which could provide significant productivity gains for 
STS operations as well as potential applications for Space Station. The 
Artificial Intelligence Lab currently has in use a number of state-of-the-art 
hardware and software systems, including: 6 Symbolics computers, 1 LMI 
computer, a VAX 11/780, a Hewlett Packard 9000, numerous personal 
computer systems, DECTalk speech synthesis systems, Votan and Kurzweil 
voice recognition systems, the Knowledge Engineering Environment (KEE) 
expert system building tool, the Language Craft natural language tool, the 
0PS5 + expert system tool, and the Automated Reasoning Tool (ART) for 
building expert systems. These tools have been used in the development of a 


■■^-'*r3^ ■ 


' number of functioning prototype expert systems. Some of the prototype 

expert systems are applicable to STS operations, including: 

Navigation Expert-- NAV/EX 

An expert system which emulates the decision making process of the 

' flight controllers wno work on the high speed ground navigation console 

during the ascent and entry phases of Shuttle missions. Currently the task 

requires three people who monitor the tracking data from 1 to 3 radar 

stations, control the operation of the High Speed Trajectory Determinator 

(HSTD), monitor the comparison between the onboara and ground 

"v navigation systems, and provide status info'-mation to the Flight Dynamics 

-^ Officer. The expert system monitors the same radar data and the output 

f from the HSTD. It will warn the operator of current or impending problems 

I with the data and will recommend potential actions. This system has the 

capability to reduce manpower requirements from three people per shift to 

4 one. 

MCC Software Status Expert System - MCCSSES 

!^- An expert system which emulates the function of the Printer 

.^ Controller in the MCC. This flight controller monitors the on-line printer for 

* error status messages during all flights and simulations. Virtu-ally all the 

'I software and hardware in the MCC report error or status information to this 

» " printer. The printer controller scans this printout in real time for significant 

•I information and the reports it to other flight controllers, primarily the 

f Computer Supervisor. This expert system could potentially eliminate this job 

V- and also provide extended capability for error detection, analysis, and 

Expert System for the Flight Analysis System - ESFAS 

An expert system which acts as an intelligent front end to the Flight 
Analysis System (FAS), a set of computer programs used to design many of the 
Shuttle missions. This system will allow less highly trained users to make use 
of the FAS for mission design and will provide a friendlier, more powerful 
interface for experienced users moving on to Space Station projects. 

MCC Workstation On-Orbit Navigation 

An expert system which will provide advice and control for on-orbit 
navigation functions. 

Navigation Console Shift Scheduling 

An expert system to aid in the complicated task of scheduling teams 
to work on the navigation console. 

Most of these expert system prototypes take full advantage of the 

^ t advanced software environments available on the LISP machines usinn .Sighly 

^7 visual interfaces and mouse oriented interaction, as well as speecii synthesis. 

Expert systems applications in more traditional languages arid computers are 

"l^ also being explored, including development of our own expert system 

i language based in C. Use of parallel processing for expert systems is also 



"jj,>^:^;%v'\ ■' • 

being researched. Along with the expert systems, the Al Section has ongoing 
work in speech recognition and natural language systems a well. 



* m^ 

.2- **^'.»8l>f>>^^~/ 



John S. Mjiaford 
Gregory B^ Bennett 

McDonnell Douglas Technical Services Company, 
Houstou, Texas 


The Space Station Program will incorporate analysis of operations 
W- constrain tc and considerations in the early design phases to avoid the need 

for later modifications to the Space Station for operations. This paper 
discusses, from a qualitative perspective, the application of modern tools 
and administrative techniques to minimize the cost of performing effective 
^ orbital operations planning and design analysis in the preliminary design 

* phase of the Space Station Program. 


Tools and techniques discussed include: approach for rigorous analysis of 
operations functions, use of the resources of a large computer network, and 
providing for efficient research and access to information. 


NASA directed that operations planning be done early in the Space Station 
program, as part of the definition and preliminary design activities. Our 
goal in Oribital Operations Planning is to avoid designing in errors that 
would later have u> be corrected, or would reduce the efficiency of 
operations due to work-around solutions. 

Because of the limited availability of analytic^], modeling tools and the 
uniqueness of each new manned space flight program, operations planning is 
the must labor-intensive techr>ical analysis done in the design of manned 
space siystems. Consequently, orbital operations planning is & prime target 
for productivity improvement. 

The buic flow of the task of operation i planning, shown in Figure 1, 
follows the method generally used for drvelopmcnt of operations plans for 
manned spaceflight. Scenarios of the operations, a reference hierarchy of 
the functions accomplished by these operation^ and resulting requirements 





Figure 1: Tisk Flow for Operatloiu Planning 

• Requircmcntt 
(Miuion, Opt, 

• Rcf. Config. 

• Opcritioni PlMtt 

• NSTS Capabilii 







Scenariot and Function liit / 
are developed interactivelv/ 

Requirement! Change Kequriit 

and Modify 
-VJ Opcrationi 

Evaluation Criteria 

• Crew Productivity 

• Impact on NSTS 

• Autonoinoui Capability 

• Commonality 

• Automation Cotu 

• Growth Flexibility 
a Crew Safety 




Requirement! Change Reque!ti and De»ign Revijio 

are developed in a iterative procedure, with each activity feeding back 
into the othert. 

The operation* planning procedure ix coupled with the process of 
preliminary design by proriding assesamenta of the designii to the design 
organizations, and feeding synthesized pro(*ucts back into the ba^ic 
reference data. .Thilc this approach ensures that operations consiJcrations 
will be included in the early phases of the design activity, the 
labor-intensive nature of standard methods of operations planning make the 
process especially challenging for productive application in a broad-scoped 
program, such as the Space Station. As the number of subsystems and design 
options for each subsystem increase,i, team coordination and information 
gathering tiemand more of each engineer's time. Since every technical 
discipline has to address operations consid4:rations, the operations 
engineers must exchange information with each design group. So, the 
effects of the communicatioi^ overhead costs are amplified for operations 
planners, who spend the largest amount of time communicatirg with other 

The keys to productivity improvement in operations planning are found in 
the implementation of these communication activities. Figure 2 summarizes 
the methods available for streamlining the analysis of the operation* of a 
large number of interrelated systems and communicating with design 
organizations to accomplish these analyses. 






Figure 2: Methods for Increasiug Productivity in Space Station 
Operations Planning 

• Organized Analysis of Operations Functions 

• Efficient Access to Information 

• Use of Technical and Management Information System 

• Standardization of Formats and Outlines 


By using an organized approach to the definition of operations analyses, 
the operations planner maximizes use of previous work done in the same 
subject areas. His efforts are concentrated on analyses most beneficial to 
the current program^ Space *!!tatlon Definition and Preliminary £>esign 
contracts divide the work into several work packages, so each design team 
is concentra*^ing their efforts on a specific subset of the total systems of 
the >>;ace Station. This situation requires an especially organized 
approach, with operations and design analyses organized so that information 
can be found quickly by the operations planner, by the reviewer, and by the 
system designer. 

Prior to thi* design study, operations analyses were organized in a mixture 
of systems, subsystems, and operational functions. The result of the lack 
of imiformity in this type of organization of the material is that a 
disproportionate amount of tine will be spent searching for data in 
comparison to the time spen<-: using the data. 

By matrixing the operations functions with the systems involved in 
performance of each function (see Figure 3), we can eliminate a confusion 
which historically arises in operations planning. Previously, functions 
and systems were mixed in this type of analysis, so that the general 
relatioiiships betweeo operations and the systems used in each operation 
were obscured. 

Each function is analyzed to determine which systems are involved, and how 
each system is used is accomplishing that function. From this analysis, 
the impact of the operations function on each system's design requirements 
can be determined. The tocal impact of all operations considerations on an 
individual system's design can be found by summing the results of all the 
functioned analyses. 


*tj *!i*W%* 


• , Mix *" ■* ■*! -* 

Figure 3: System Affected by Operations Functioru 

For efficiency, only major design drivers are considered da-ing preliminaiy 
design. The resultant operations scenarios are incomplete; however, since 
the scenarios are developed and modified iterativcly with the system 
designs, they provide a foundation for later phases of the program. 
Carrying the rarly work into the detailed design phase avoids additional 
costs to perform the same analyses again. This process provides the 
foundatif • for highly productive development of further documents when the 
progr^.m has reached sufficient maturity to began detailed planning of the 
ster,-l;>-stcp procedures. 

Providing a matrix of operations functions and the subsystems involved in 
accomplishing each function enables all users to find the data efficiently 
and conveniently. The function heirarchy is based on current experience 
and practices in the Space Shuttle program, extended to encompass the 


Figure 4: Program Genealogy Chart 










unique features of the Space Station. This arrangement closely {(.i^rallels 
the organization of NASA personnel and provides a familiar form that allows 
reviewers to find what they are looking for quickly. Operations planners 
can assess the completeness of their work and identify the trend of system 
impacts. And subsystem designers are provided with a quick look at the 
bottom line of how operations considerations place requirements on their 


Operations planning engineers spend the largest share of their time finding 
the information needed to do their work. At the beginning of the program, 
a training curriculum should be developed that teaches the engineers what 
work htu been done before, and where to find the applicable documentation. 

The training should include sufficient details on past programs for the 
engineer to assess the applicability of the past work to the current 
problem. For example, the basic hierarchy of operations functions of the 
Space Station have not changed significantly during the past twenty years, 
though the details, emphasis, and methods of implementation of these 
functions have changed dramatically. 

Genealogical histories are most useful to training engineers to find and 
assesM information quickly; the genealogy of the program, and the geneaology 
of the available documentation. By representing the interrelationships 
of programs and documents, by objective and date, a better understanding 

of the current program's history and applicable documentation is 
available. Eximples for these arc shown in Figures 4 and 5. 




Hgure 5: Genealogy of Applicable Dociuatntation 

Space Sution NeuU, 
Attributea, and Arzhitectural 
OpUona Stmliw (April 1913) 


NASA/JSC CoBccpcual Deai|B and 

Evaluation of Selected 

Space Station Concepu (Dec 19(1) 

Space Station Opcratlooi 
Plan. Appendix B: 
Orbital Operation! Plan 
(Dec 1*85) 

Space ! tatlon Operatloni 

Plan. DR-07. MDAC 

(Dec 1913) I 

(Fall 1913) 


Space Sutioi Flight 
Operation! Plan 
(Fall 1915) 

Space .Station Definition and 
Prelijilnary Dcilga Study. 
Requeit for Proposal, 
Appendix C-3: 
Operation! Requlrcmenta 

Space Station Flight Operation! 
Description a£d Deilgn Criteria, 
NASA. JSC-20433 (April I9«5) 

Space Station Sjiteni 
Operational RequlremenU 
(Dec I9S3) 

Space SutloD Gperatloui Plan 
NASA. JSC-19946 (Sept 19«4) 

Spaca Station Orbital Operations 
Plan. MOTSCO-HAD (July 19K) 

Space Station Program 
Desrripticn Document. Book 6. 
Sr.tion 10: Operation! 
Rrquireraeatj (Sept 1983^ 

Spacp Station Program 
Deicnption Document, 
Book 6 Syitcm Operatloni 
(.Sept 1983) 


Sutlon Program 
ption Document, 
Station Task 
(Dec 82) 

Space SutiOQ Program 
Description Docun«ent 
(Aug 19S3) 


Opcratioai C^ 

Besides an elaborate computerized communications and data storage network, 
a well-organized central libary provdics a cost-effective productivity tool 
avidlable to a large program. The cost of implcmertation of a library is 
usually one research libararian plus the fl(Kir space for the facilities. 
Some of the costs of facilities for storing documentation will be recovered 
because the presence of the central libi-ary reduces the need for similiar 
furniture to store several duplicate copies of the same documentation at 
each engineer's work site. The increased productivity from having a 
library available can run between 25% and 90% of each engineer's time, 
depending on the program and the er mincer's specific work assignment. 

Beyond provision of a library and the necessary tools to use the library 
efficiently, electronic tools can hr used to o'ecrease the time spent 
discovering and gathering information. Methods for using these electronic 
tools arc discussed below under Use of the Technical and Management 
Information System. 




^|i^;.--^ft- •■.'^•«' ^■-•■- -. -• ' tA. 


Figure 6: Ck)mpariso]i of Older Methods to TMIS Toolt 
Previous Method Replaced by 

• Hand-written text • Word processing 
and typewriter 

• Overnight express • Electronic mail 

• File cabinets and • Electronic storage 
bookshelves and retrieval 

• Drafting boards • Computer graphics 

• Travel • Video conference 


Engineering labor costs at least 100 times as much as a computer's time. 
(The exact ratio of the ost of adding an engineer to a project compared to 
the cost of serving an additional computer terminal or adding a desktop 
computer will vary depending on the tools and computer configurations used 
in each program.) So, computer systems should be designed so that the 
machine is always waiting for the man, and not the reverse. 

NASA specified that all products of the Space Station definition and 
preliminary design study would be "delivered electronically to the maximum 
extent possible". The Technical and Management Information System (TMIS) 
which resulted from this direction has provided us with an unprecedented 
kit of productivity tools which have proved their worth in areas beyond the 
simple electronic delivery of documentation. 

Figure 6 compares tools prevously used in operations planning to those 
available from the TMIS. 

Electronic mail has replaced most of the daily technical coordination that 
otherwise would be done using typed memos. An immediate gain in 
productivity usually results from having engineers typing instead of 
writing by hand. When revisions and editing are performed by the engineer 
himself, without secretarial assistance, he avoids errors that result from 
miscommunication and from the secretary's interpretation of Lis cryptic 


■ .-.■^^iT^: . 

:V V---*\>'.'- ^-. ■- ' t^' 

Figure 7: Summary of Data Base Applicationa 

• Detailed Task Schedules and All Schedule Reports 

• Budget and Time Tracking 

• Requirements Analysis 

• Crew Time, Skills, and Training Requirements 

• Resource Analysis by System and End Item 

• Operations Functions Hierarchy 

• Reference Docimientation Listing and Evaluation 

• Tracking Training of Enjin&ering Personnel 

• Tracking Task Products and Documentation 

Use of electronic mail also provides a machine-readable form for the 
contents of each message. The text and technical analyses contained in the 
day-to-day engineering communications can be merged and formatted into 
deliverable documentation with minimal additional work. Having tsxt and 
graphics in machine-readable form eliminates the extensive duplication of 
effort necessary to produce final documentation from paper copies of the 
engineering work. 

A key feature of the implementation of this philosphy is having the TMIS 
interface electronically with the word processing facilities at each site. 
Products of these daily activities are in near-final form when they are 
transmitted to the word processor for formatting into fo^m^d documentation; 
and they c^ul be transmitted with no additional keystroking required. This 
procedure can reduce the man-hours required to produce a document by an 
order of magritude. 

Data Base Features 

The centralized TMIS also provides a data base system which is applied to 
enhance productivity in the areas summarized in Figure 7. 

The data bases used in daily technical analyses contain relations of all 
mission, operations, and system requirements, the function breakdown, the 
operations scenarios, and design assessments. These data bases are used to 
create figures for final deliverable documentation from the technical 
analyses presented to system design groups. 




\rtV '>*JE'. ..■*»«,. 


Data bases also used to reduce the labor overhead in management of the 
task. One database contains an extremely detailed master schedule for the 
task from which reports are extracted so all the engineers are aware of 
short-term and long-term milestones. Other reports from th« detailed 
schedule data base keep the customer informed of when to expect certain 
products from the sf^udy, and still others become inputs to program-level 
master schedules used to manage the whole team's progress and direction. 

Tools can be developed as they're needed 

The need for new software tools arises continuously in the R&D environment. 
The availability of a computer and a high-order language provides the 
engineers with the basic resources they need to generate these new software 
trx)l8 when the need for them is identified. If at least one of the 
engineer* in each group is proficient computer programming, analyses that 
would have taken weeks of drudgery with a hand calculator can be dispatched 
with a smaU investment in programming time. 

For example, in the analysis of the Space Station personnel transportation 
needs, a few hours of one engineer's time produced a program to model the 
status of Space Station and Space Shuttle crews and vehicles. The program 
provides a top-level look at the transportation scenarios for any set of 
input parameters, such as variations on the Station crew sire, maximum stay 
time on orbit for a crewman, maximum number of people on board a Space 
Shuttle Or biter, skills required on beard at a specific time, and so forth. 
Without the TMIS resources, this analysis would have taken longer to 
produce less rigorous answers. Use of the TMIS also produced 
documentation-quality output, eliminating artists' time and providing 
engineers with data in an easily understood form. 


Standardized formats and outlines bltc used to speed compilation and access 
to information. Standard froms for each bookkeeping job (such as recording 
applicable requirements and system effectivity of each operations function) 
alert designers to the operations concerns that have a significjmt impact 
on the design of their individual subsytems. 

Standardized outlines also will reduce the amount of time required to 
review and synthesize the inidividual inputs of each work package. In the 
Space Station studies, all work packages depend on NASA to assemble these 
individual inputs and republish them for use by the entire program. The 
productivity gains from an efficient synthesis process count twice, since a 
timely turn-around of these data will keep all the work packages focussed 
on a common baseline and reduce duplications of effort. 

These standard formats work well with the data base management features of 
the TMIS. Relations between requiremcnbi, functions, operations scenarios, 
subsytems, and sections of final documentation are preserved in data bases. 
Reports from these data bases show up-to-<late relationships betwen the 
data, and immediately identify missing information. 




Figure 8: Enchancements to Productivity Tools and 
Techniques Currently Being Developed 

• Database of graphics for building figures 

• Automation of including grahics in documents 

• Increase number of network terminals and 
desktop computers 

• Flexibility so each engineer can use the 
machines and tools he's familiar with 
and still work with a mainframe-based 

• Get all ( .)Cumentation on line, including 
historical libraries 

• Increasing user-friendliness of engineers' 
tools to reduce training time and efficiency 
of use 

• Standardization of computers used 

• Multiple-purpose tools^ make tools friendly 
enough to enccjxage their use in early 
design phases 

• Common data bases - system designs, 
simulators, graphics systems, visual 
generators, logistics planning systems, 
failure analysis programs, training plans, 
manifesting systems, stowage tracking 


Experience with these techniques for improving engineers' productivity 
leads to come reasonable goals for building on the available tools. These 
goals are listed in Figure 8. Each of these goals is at some stage of 
evaluation and implementation in the Space Station program. 



L/ • ---z-^SS^- ■■:■■■ . 


John Manford is an engineer in the Space Station Orbital Operations group 
at the McDonnell Douglas Technical Services Company in Houston, Texas. He 
receired his Bachelor of Science degree in Mechanical Technology from the 
University of Houston in 1981. His work experience prior to joining MDTSCX) 
in March 198S included product engineering for an oil field wire-line 
service company. 

Oreg Bennett is the Task Leader in the Space Station Orbital Operations 
group at the McDonnell Douglas Technical Services Company in Houston, 
Texas. He received his Bachelor of Science in Aeronautical and 
Astronautlcal Engineering from the University of Illinois in 1973. His 12 
years of aerospace engijieering experience Includes exploratory design of 
commercial aircraft, simulator development and operations, and space flight 
operations planning. 


-1 393 



•.1fvj», -"^ ' 

\ ? • 



Peter R. Ahlf 
McDonnell Douglas Technical Services Company 
Houston Astronautics Divisio.n 


As the Space Transportation System (STS) enters its operational 
era, the need for standardized, automated, mission planning computer 
tools has become apparent. In order to support an increased flight rate 
without a corresponding increase in manpower, quicker and more efficient 
methods are needed to perform standard tasks. Also, the problem of losing 
experienced personnel through attrition creates the need to retain their 
knowledge and expertise even after they are gone. 

To help solve both of these problems, the Shuttle Trajectory and 
Launch Window Expert System (STALEX) was developed to automate many as- 
pects of early mission planning for space shuttle missions carrying geo- 
synchronous communications satellites. Most of the commercial Shuttle 
missions planned for the i" t three years will carry at least one of this 
type of satellite. The a^,. .:ations of ST/>LEX include payload deployment 
scheduling, launch window analysis, orbitai trajectory determination, and 
landing opportunity selection. 


Recent history has displayed a virtual explosion in the popularity 
and visibility of artificial intelligence (AI) applications to industry. 
The AI technique receiving the most attention is that of "expert systems". 
An expert system is a computer program which duplicates the decision pro- 
cess which a human expert would use to solve a complex problem. Most 
expert systems are written in relati«/ely new AI programming languages 
(e. g. LISP). These languages are being supported by a new generation of 
computers which are designed with emphasis on the machine-user interface. 

AI oriented languages are based on a different concept than their 
predecessors. While languages such a"^ FORTRAN, PASCAL, and Ada are de- 
signed to manipulate numbers, languag. such as LISP, LOGO, and PROLOG are 
designed to manipulate characters or symbols. This approach, known as sym- 
bolic processing, is conducive to the modeling of decision logic; a process 
known as knowledge representation. 


'■'*'* '^^SS^ 

k »'^^ -"^*. • ■•►^ ■ 


STALLX is an expert system that is used in the pre-mission planning 
frr Space Transportation System (STS) shuttle flights. It was developed on 
a SYMBOLICS 3600 computer and is written in a combination of LISP and 

This paper describes, in general, how the use of expert systems and 
the latest machine-user interfaces can i^iprove productivity. Specifically, 
the use of STALEX in the early phases of Space Shuttle mission planning is 
described. To highlight, the benefits of expert systems, a close compari- 
son between STALEX and the old analysis tools will be made. 


The STALEX is used for flight design on Space Shuttle missions 
which carry geos>,ichronous communications satellites as their payload. For 
this type of mission, the STALEX performs two major tasks: launch window 
analysis and scheduling of when the payloads are released (deployed) from 
the Shuttle. 

The launch window is the range of time during the launch day that 
launch may occur. Satellites impose constraints on the launch time due to 
their requirements on the inertia! orientation of the shuttle parking orbit 
at the time they are deployed. Because the shuttle orbit does not remain 
inertially fixed throughout the mission (it is rotating due to grai/i tation- 
al perturbations arising from the oblate shape of the earth) the launch 
window requirements for a satellite are also dependent upon how long into 
the mission they are deployed. Each satellite is scheduled for one pi ime 
and one backup deployment opportunity. Therefore, each satellite imposes 
two restrictions on the launch window; one corresponding to its prime de- 
ployment opportunity, and one corresponding to its backup opportunity. 

There are also restrictions on when during the mission a satellite 
may be deployed. Orbital mechanics involved for the satellite to reach 
the desired final geostationary orbit limit deployments to occur once every 
90 minutes as the Shuttle passes over the equator. Additional payload re- 
quirements limit deployments to occur on only those equatorial crossings 
which fall within certain longitude bands. Finally, the times of the sat- 
ellite deployments must also be compatible with the astronaut activity 
timeline. A few examples of these considerations include limits on the 
number of events scheduled for any one crew day, the unavailability of 
those revolutions during the crew sleep periods, and requirements for the 
amount of tinie between scheduled deployment opportunities. The mission 
planner must determine a deployment sequence (a prime and a beckup oppor- 
tunity for each satellite) which optimizes the criteria for the times of 
deployment and the resulting launch window. 

The solution of this problem involves a '"'^-bination of straight- 
forward numerical analysis and the application of expertise gained from 
experience. The conversion of the payload requirements for deployment on 
a given orbit to a launch window requirement is an unambiguous mathematic 
calculation. On the other hand, the guidelines for scheduling deployments 
from a crew timeline standpoint, or more importantly, the criteria for 



selecting the best deployment sequence from a large set of possible solu- 
tions, are not hard jnd fast, and most a^e not officially documented. In 
most cases they arp simply rules of thumb that are learned from experience. 
The STALEX combines models of these rulos (a process known as knowledge 
representation) with the numerical calculations needed to determine deploy- 
ment opportunities and launch winaows. Its goal is to allow ease of use, 
to perform the task quickly, anj to consistently choose the same solution 
as the experienced mission planner. 


The concept of the "expert system" involves a basic divergence from 
earlier approaches to computer programming. In conventional programming, 
the computer is used to aid the user in his decision making/problem solving 
by speeding up numerical calculations. This method, shown in Figure 1, in- 
volves the expert in an iterative process of calculation and decision making 
leading to a solution. 






Figure 1 

An expert system models the knowledge used by the expert in the 
decision process, thereby removing him from the loop (Figure 2). 


The expert system appro 
First, if the knowledge of many 
system shouia Lneorctic^l iy out 
pert system will operate faster 
boredom, forgetfulness , or fati 
able to industry and the space 
performs as well as the human e 
a problem is "captured" and wi 1 
around. The aerospace business 
nel turnaround, and the cost of 

Figure 2 

ach holds benefits from three standpoints. 

experts is accurately represented, such a 
perform any single expert. Second, the ex- 

and wi "nout the human drawbacks such as 
gue. A third benefit is especially applic- 
program. If an expert system successfully 
xpe*-t, tne knowledge and capability to solve 
1 not disappear when the expert is no longer 

is known to be dynamic in terms of person- 

T>nlacing expertise is thus substantial. 






These concepts were used in the development of STALtX with the 
key goal being that of autom a tion . 


The task for the flight designer is to select a satellite deploy- 
ment sequence which meets two criteria. First, the schediilad deployments 
must meet the crew timeline guidelines. Second, deployment sequence must 
result in an acceptable launch window. These two conditions do not always 
result in the same deployment sequence. Therefore, many potential solu- 
tions must be identified and evaluated to determine the one which best 
meets the requirements. 

Previously, five separate programs (the Payload Launch Window pro- 
gram family) were used to aid in this analysis. Two of these were used to 
enter and edit data files which define the payload constraints on the 
shuttle orbit orientation at the time of deployment. Having constructed 
these files, the following steps were used to determine the best solution. 

1 - Determine available deployment orbits fr.aC meet nayload constraints 

by running the Payload Deployment .Opportunities Launch Window 
(PLDOLW) program once for each satellite 

2 - Select a deployment sequence that meets crew timeline constraints 

(a prime and backup deployment opportunity for c^ach payload) from 
the PLDOLW list of availabie orbits 

3 - Calculate the launch window corresponding to the deploymtftit se- 

quence. For the planned launch date, this i: done by running the 

Composite Launch Window Plotting (CLWPLT) program. For a range of 

dates around the planned launch date, this is done by running the 
Payload Launch Winduw Plotting (PLWPLT) program 

4 - If the launch window is ur, acceptable, or if it may be improved by 

changing the deployment sequence, select a new sequence and calcu- 
late the new launch window by running CLWPLT and PLWPLT over again. 

This process was highly iterative. One program had to be run be- 
fore the user could select a potential solution, then another program had 
to be rut. to find if the solution was acceptable. The second and third 
steps had to be repeated numerous times until the flight designer was con- 
fident of having found the best solution. 

The major drawback was the lack of automation in these programs. 
As described before, the final launch window is the resu''t of combining t^e 
launch window restrictions derived from several constraints. In running 
the CLWPLT and PLWPLT programs, the user would specify a constraint, and 
the program would calculate the corresponding launch window. The user 
would then specify the next constraint, and again the program would give 
the resulting launch window. The CLWPLT program would calculate the inter- 
section, or composite, of those previously entered constraints specified by 
the user. This process is very labor intensive, requiring the user to bf 
constantly entering data and waiting for computer to perform its calculations, 







An expert system requires a ^^ery different approach. STALEX 
combines all of the algorithms required for the numerical calculations 
(steps 1 and 3) with a model of the knowledge used to select and evaluate 
potential solutions [ll. There are two types of knowledge that must be 
modelled; the methods of generating possible solutions, and the guide- 
lines tor evaluating the solutions. STALEX uses a converging process to 
generate and evaluate a potential solution set. Beginning with the realm 
of all possible deployment timelines, a large number are eliminated and 
the remainder are evaluated. STALEX models Ihe human logic needed to do 
this by using filter s and rules . 

Filters are functions which eliminate large numbers of potential 
solutions by using certain terminal criteria. Failure of this criteria 
implies that a solution is totally unacceptable. Rules may also elimin- 
ate solutions, but they model the less stringent guidelines needed to 
distinguish an optimal solution from many acceptable solution. 

With the computer performing the second step, the human expert 
is removed from the loop and the entire analysis is automated. This 
approach lends itseli to a more graceful style of data entry. Because 
the program now performs the entire analysis all of the top level data 
must be input prior to its execution. .The user enters data once at the 
beginning instead of constantly during the program execution. 


An aM too common occurrence in the space program is an unex- 
pected problem causing the definition of a shuttle mission to change. An 
engine problem can cause several launch dates to slip, or satellite manu- 
facturing delays can cause the payloads on several missions to be remani- 
fested. A typical shuttle mission may see two launch date cl' mges and 
three payload remanifests before actually flyinc. In these cases, the 
launch window and deploy sequence must be recalculated. Reacting to such 
changes quickly and accurately is imperative in order to maintain a rea- 
sonably stable STS mission schedule. The STALEX accommodates this by 
making use of previous executions of the program to speed up the wor'' 
needed for mission redesign. 

The original PLW program family used one mechanism to save time 
on repeated executions of he programs. By building files of payload 
constraints, this data was entered only once. The STALEX expands upon 
this technique by allowing the permanent storage of an entire mission 
definition. All of the data needed to determine the launch window and 
deployment sequence for a mission can be stored in a data base. Should 
a change occur on a mission, that mission data is read into the program 
from the data base (rather than re-entered from the keyboard) the change 
made to the definition, and the analysis performed for the new mission 







Conventional computers communicate with humans by displaying 
characters or graphics on a display or screen. Humans, in turn, com- 
municate with computers through the use of a keyboard, light pen, or 
mouse. The method of interaction between the human and computer is 
called the user interface. The levels of sophistication in computer 
user interfaces range from the single number display of the hand held 
calculator, to the advanced CAD/CAM computers which can display three 
dimensional representations of objects which the user can move or rotate 
with the touch of a light pen. The capabilities and ease of use of any 
program are dependen. upon the user interface of the host computer. 

The ciginal launch window design programs were developed on a 
Hewlett Packard (HP) 9825 desk top calculator. The term calcul.^tor is 
misleading here: the system included floppy disk drives, a pen plotter, 
and a dot matrix printer. Its major disadvantage was that the "screen" 
displayed only one line, not unlike a pocket calculator. This forced 
the programs to be written in a "prompt/response" style. The single 
line displays a prompt, requesting a value for a variable or a choice 
for which program operation is to be performed next, and the user re- 
sponds. The drawbacks of this type of user interface include: 

The user can only view a bare minimum of information at any one 

The number of characters on the single line pr'event all but the 

shortest prompts or messages 
The values of program variables are hidden from the user because 

the single display ^ine is erased with each new prompt 
The user has minirrial control over the flow of the program processes; 

the prompt/response sequence must be followed identically each time 

The STALEX is hosted on a SYMBOLICS 3600 computer. This computer, 
known as a LISP machine, is one of a new generation of computers designed 
for AI applications and featuring advanced user interface capabilities. 
The system includes a large high resolution screen, keyboard, and mouse. 
The user interface is based on the window concept [2], A window is a con- 
figuration of the computer terminal screen. Each window is divided into 
smaller sections called "panes". The STALEX makes use of two types of 
panes; output panes and menu panes. Output panes are sections of the 
screen where graphs and dynamic text information is displayed. Menu panes 
in turn consist of three types [3]: 

Variable value editing menus which display and allow editing of 

program variables 
Cormiand menus which allow control of program processes (STALEX 

displays these ir inverse video) 
Temporary menus which appear following selection from a command 

menu in order to re-f'iiie trie choice 

The STALEX consists of six windows. The top level window is 
called the Mission Definition Window (Figure 3) and is displayed upon 
program initialization. This window is designed to allow entry, editing, 
and viewing of all the basic mission definition data. This data may be 



^\ „ i • Ski > "' 
4 -.^-.X..-^- 


I imn I'MU pi fill 

r **\-i\if I *• t"-i ■•" »>^ 

M*Mi4t-l •>••• ''^^ r41l<>4l>-liA t «I>«M 




I.MMH ttfC (*'«/•} 
••«C VTtlM . ■ ■ ■ 

■«••(■ H» i*ftri**ai 

■ItllCI fftli|«LC . 

•tH r»H iCni. ■ 
•WHIl •# CM« 
IITIV «Mll . . 
k«l»t« «f titt Kit 



a t 1 

t. 4 




LAtl!,^ Myt^- 

I.*M1M f'Tt I . 

NT f L4M1M 1 
CMSIftAIlT 1 . 

L*M(«« «IIf t 

«Cr f L&MIM t 

CMwruiir t . . 

■tuiai • . 
• IM tire « 

CMftTuiar « 

■cr tf LuaiM f 

13 3 4 1 



• a i4jiM 
Aa w-11 

UO tOw itOA 
KAQ 0«A eZA 

UO cow t^om 

lUkJ D«/l ZZA 

Mi M-11 


m A»-ic 
UO cow Hom 



latcmta A*MI> (■ ■*)> iM.t 
laWIIIH fCIIICC (■ «t>t !««.' 
IMltUVlM I ».«1 

■MC WBMi : 2 

«C1 9* *•■ IMC . I lilttSiM 

itMIIMC •* ft atMi t'l.ll 

■«! ar acte aaae ■ •iit39ii4 
caaiiiaai ar aic laaei -it.if 


iwrci tfAtC 

laxcria* tfwt 
•CKL9T ritrcicitc 

caaaii*)«t ifrc 

JC 0<1' 
I > 3 4 

tAMt CCOS-7 

taJlCtlM fC AlCMkC 

•cpiar ricrcuacr - « i 3 4 

c»atTa«i«t tVM 

tt£y TlntLlhf BEBI 

n.aa a^ci 
rai »fla icaii 

rit PI4 cia 

'■B pt« icaii 

paa P4« »Tiia 

.'»r »r« Mtia 

f— pan caa ■ 

aii7iti0 a 

iriJiMis a 





4.13i3«>l • 





t f itvaumc tthaf- o t. 7 ] 

tlT tata iTPt 
C4tf illll 
ICIT Llllf 

faa aua fpt 

Cfc«T lltit 

laa aaia tv^c 

C4IT llllt. 
BOI Llllt 

L IaCmuP 
»*»< aAc«u» 


iwtMc aAcu* 



« pf La««i>va< ■««» 

HAMO lorT 

13 3 

IIT ■•■8 tfPt 
C4II illlT 
«(«r Llllt 

CM a««* tvM 

C»I1 Ll«11 
flCtI Llll*- 

sia iAa* fvc 
CMC (.lait 
wcat LI4II 

I aftcwj* 
I m. 

I «nMc aACKu* 

1 ML 
: ML 


1 ML 

' Ml 




urrcB iiAVC . 


tCPLav ricpcific? 

caatTiAiiit ITPC . 


*ae ocac 

13 3 4 


lAac. . . . 

u#rca aTAac . . . 

1I3CC11» \1P% 

atPLaT ricritiacf 




f a 9 « 

rrc . I HAMO •on 

f ar iiaaiTiMiC a«iafl. 

tiT aAaa UPt 

cur IIKIT. . 

•car iiair. 
caa aiaa rvPt 

CkBT Liaii 

9la >»■" TTPC 
C««T LtBU . 

vCiT 1 iiir 


* ap LaaaiTuac aaaaii 
18T aaaa upt 

CiaT LIIIT. . 

•car Liiir. 
laa a«a< itpt 
CBtT Liiir. . 


saa a«aa irn 
tfcBT Liaii 
fcar l:iit 









Figure J 

■ >l 1 Mi'\-iH'I>"\'* 

fit* ifim 

Pi.*% TlfLIi CMWi^E LftUrtCH ytrtOOu GMPfu 
BTBiT aar: \9m. 
caa Bar- 13M. 
aiBB PCI rtCi 1- 

nc» »CI tfcMLi ' 

taa aiit tiBC icaci CST 
aaiB caaraBirci vu no 

OCOBM a*OKl» BCV !■ 

■ •I PMa« arv ta 
Bcoa-x aAoujp rcv 4b 
ocoa-i ma* acv «i 

DCOl-a BACKUP lev 4B 

' B t «a aa 
cac 4 ti aa.aa 
lac i It 14 41 



laat naa i3m iim 14M ^ x 

oco*-i p*mm. arv • 
onia-i lACKi^ acv t* 

aeo»-> aA£au» w 4b 
oena-B pmmc arv aa 
flcoa-t aAoit^ arv t% 
EDW B I 4a-aB 
Ksc 4 zk aa «B 

KJC 4 12 l« 41 





—f— * — t — 1 — t — ( — I — t — ♦ — i — i — t—*— 

I fcaa raa aM 




entered manually or automatically loaded from the mission data base. Two 
other windows are used to enter data into the permanent payload data base, 
and the remaining three windows display the launch window and deployment 
sequence solutions. Figure 4 shows an example of the day-of-launch 
launch window display. 

Rather than a prompt/response flow of control, the STALEX is a 
' menu driven program. To change the value of a program variable, the user 

<i positions the mouse over the current value, clicks the mouse, and enters 

'I the new value from the keyboard. Other menus allow program processes to 

; be invoked by again clicking the mouse on the desired action. This type 

J of user interface has many advantages over the prompt/response concept: 

All of the basic inputs are logically formatted and visable on 
" one screen 

Using the mouse to make selections greatly speeds up the editing 
process and reduces the chance of typing errors 
• . All inputs are predefined and are drawn from requirements docu- 

;, ments; no design work is needed to calculate the input data 

I Erroneous inputs are trapped at a high level and re-entry is 

;| permitted 

F A large set of default data is available and automatically loaded 

fif the user fails to provide all of the necessary inputs 
.^ The user interface is highly interactive 

■' ^' 

The STALEX graphics and graph enhancement capabilities provide 
additional time savings for the engineer user. A common problem for all 
engineers is to effectively communicate the result of an analysis. This 
is especially true for the flight designer who must present the launch 
window analysis to several NASA mission planning review boards such as 
the Flight Operations Panel (FOP), the Cargo Integration Review Board 
(CIR), and the Mission Integration Control Board (MICB). The launch 
window problem is not easily visualized, and the only method of commun- 
icating the results of this analysis is through effective graphics. 

The PLW programs plot results directly onto paper. For a -typical 
^'- launch window, this would take approximately ten minutes. If a mistake 

; was made, the process had to be completely repeated. The STALEX dis- 

plays the graphs on the computer terminal screen, in less than a second. 
The graph can be enhanced or replotted in a negligible amount of time. 
When the user is satisfied, the graph can then be "dumped" to a graphics 
printer. The ability to use interactive computer graphics is dependent 
upon the capabilities of the computer, not the inherent capabilities of 
the program. The STALEX also includes such graph enhancement features as 

4 automatic region shading, labeling, grid and line drawing, and arrow 

* drawing. These features make heavy use of the mouse to provide a natural 

* and highly interactive graph enhancement capability. 


■.v.v l'•'-«L^''-, 



required to pe 
tnent sequence 
Figure 5 compa 
tasks. Figure 
combination of 
amount of desi 
values in Figu 
able satellite 

s 5 and 6 provide a represent;? five cor"parison of the time 
rform several tasks relating to launch window and deploy- 
determination using the PLW program family and the STALEX. 
res the time required for data entry and graph production 
6 compares the larger analysis tasks which represent a 
smaller tasks including those shown in Figure 5. The 
gn time is r Jated to the complexity of the mission. The 
re 6 are representative of a mission carrying three deploy- 



PLW Programs 

Figure 5 



» I ..» 


t^7\ Inlllol 9»U tntrf 

t\^ H«paoU4 Oala Cnlry 

E^S Croph Ho'dcapy 

^^^ CropH CnHoncamvnt 

yXl Anslyi* D«pl*r Oppa 

PP^ PoylOQ^ Data Entry 

IPTTI Dapl«y S«q Analyala 

(^3 Lounah Wlndax Oailgn (ln»'"0 

^3 Launch WIndaw 0»*lgn (Naw Lavineh Oala) 

Q^ Launch W1nda» Oaaign (Na» C«rga) 

cog ratal Daal0n (par mta'lan) 

PLW Program* 

Figure 6 

-'•w^pj -.• '' 


In addition to the obvious time savings the STALEX is usually 
nuDre thorough in its analysis. For a typical mission carrying three de- 
ployable satellites, the STALEX evaluates over 500 deployment sequences 
and calculates the launch windows for the best twenty solutions. Using 
the PLWPLT programs, the human expert would examine ten sequei^ces at the 
most. While STALEX discards over 90% of the solutions It evaluates as 
unacceptable, It will never overlook an acceptable solution while the 
human expert might. 


The STALEX is a prototype system. It nas been used for launch 
window design on six space shuttle missions. For three missions it has 
been used as the sole launch window design tool. It has proven the use- 
fulness and power of applying AI techniques to solving engineering 

Experience with STALEX has led to two conclusions. The first 
is that the use of Innovative computer programming techniques to automate 
design tasks can greatly improve productivity. Expert systems can reduce 
manpower. Improve quality of work, and tnake the engineers job more re- 
warding. The second conclusion is that a computer program can only be 
as powerful as the computer that runs it. The capabilities of the com- 
puter (e.g. graphics, user interface, memory, etc) limit the capabilities 
of the program. 


[1] Ahlf, Peter R. , "The Shuttle Trajectory and Launch Window Expert 
System", Proceeding s of ROBEX '85 , Instrument Society of America, 
Research Triangle Park, North Carolina, 1985, pp. 129-130. 

[2] Symbolics, Inc., "Using the Window System", VOLUME 5 USER INTERFACE 
SUPPORT , Symbolics Inc., Cambridge, Massachusetts, 1984, pp. 1-137. 

[3] Symbolics, Inc., "Window System Choice Facilities", VOLUME 5 USER 
INTERFACE SUPPOR T. Symbolics Inc., Cambridge, Massachusetts, 1984, 
pp. 1-79. 


The author received a Bachelor of Science degree in aerospace 
engineering from the University of Virginia. He has been employed by the 
Houston Astronautics Division of the McDonnell Douglas Technical Services 
Company for two and a half years. He supports the Flight Design and Dy- 
namics Division of the NASA/Johnson Space Center performing flight design 
integration and orbital trajectory design fo'' Space Shuttle missions. 






Paul Hollingshead 

McDonnell Douglas Technical Services Company 

Houston, Texas 


Preparing 30 controlled documents for each flight of the Space 
Shuttle involves 90 people at JSC in processing 200 or 300 changes to 
existing checklists, and building new procedures. A prototype system is 
discussed below which stores and displays those precise sets of 
instructions, including rationale and change history, using readily 
available hardware. In addition, 1t provides a sinfiple method of 
structuring the data in a form for easy understanding and maintenance. 

Crew Procedures Attributes 


Are the distilled result of much study and decision making 
by experts in multiple functional responsibilities. 

Have a definite process for coordinating and implementing 
the frequent changes. 

Us? standard names to refer to switches or indicators, as well 
as having plain text, and short tables of data. 

Must preserve the sequencing of steps to work. 

Refer to specific people, times, or pieces of equipment that 
need to be orchestrated. 

Comply to some standard format. 

(Notice that the above description applies not only to 
procedures for Shuttle operations, but to test procedures fOT- many 
other types of complicated equipment. A subset of tho5.e qualities also 
applies to many contractural documents and interestingly enough, to 
program source code.) 





— f 


The Problem 

,^:" More information is related to a page of procedure than just the 

".■- instr'jcticr.s "listed. There is a background of what the step does, or 

I whose inputs it was oased on. In the case of switch labels with obscure 

%". abbreviations, it's useful to have the full name shown, with a brief 

1- description of action initiated. Some explanation of the rationale for 

\ a step (or group of steps) is needed to acquaint a new crew member or 

¥;, flight controller with the steps being performed. In the case of a 

'v:, payload supplied by an outside organization, it may be useful to make a 

reference to their requirements document as the reason for a particular 
-| way of doing things. Few people have the expertise to understand the 

rationale for all guidelines given in disciplines other than their own. 
^'^ An additional volume, such as a Flight Procedures Handbook is one way of 

explaining the steps. In the heat of preparing the main procedure for 

an upcoming flight, it is inconvenient to spend the time needed to keep 

that separate documi?nt up to date. 

;. Much of the information in the checklist isn't easily collected 

L with only manual methods. Engineers specializing in a particular area 

' |: may want to see only the operation of a particular piece of equipment 

J for comparison with other books or flights. Individual crew members may 

! f want to skip steps for the other crew members and pick out the next step 

; he or she should perform. At a given point in operations, it's useful 

to know just what configuration a panel or system has been put in, and 

trace the steps that change the setup. Gathering that information by 

hand is slow and tedious. 

Paper distribution for the people involved in change 
coordination has certain hazards. A fifth generation copy, especially 

': if it was handwritten, can take extra time to interpret. In typical 

distribution systems the proposed change often spends days being copied 

./ and routed. If copies are sent out in parallel, then one person has no 

way of seeing how other members of the review team have responded. If 
the change request is routed serially, then changes can't be done 
quickly, and it can be tough to decipher just which one of four previous 
people is making any particular suggestion. 

Modifying the masters for a book, using mostly manual methods, 
and distributing copies that incorporate the approved changes can take 
weeks. S&ne portion of that preparation time is spent in ensuring that 
the procedure conforms to an established standard for combining the 
plain text, switch throws, tables of data, and computer inputs in a 
consistent manner. Editors also check to see that only approved 
nomenclature has been used in switch and talkback names. If there are 
specific numbers uscid for each step, those may need to be redone, which 
can ripple through several pages. 

For Shuttle missions the last weeks before flight are periods of 
I both intense training and the final editing and printing of the on-board 

'i:--' documents. For new refinements or problems found and solved during 

\''^- that final practice, two alternatives, neither of them attractive, 

r present themselves. Either postpone the inclusion of the change until a 

■" \ flight or two later, or go through the extra, often frantic, effort to 


M#*^. '-'-^ S ■■>■ 

get the change in the books for the next flight. Flying, or even 
practicing with an outdated checklist is to be avoided. The increased 
flight rate demands that the preparation period be trimmed down. 

The established method for coordinating changes involves many 
people, but certainly doesn't reach everyone who is interested in a 
modification. To someone who has reviewed the checklist, but doesn't 
work with it daily, the most important feature of a new edition is the 
changes since its last use. In paper documents, "change bars" indicate 
that an improvement has been made, but offer no clue as to why, or who 
requested it. Oftentimes there is a history of changes to explain why 
a sentence endi?d up that way. Some changes don't have their own 
technical justification, but are made merely to be consistent with 
similar documents. Change bars by themselves leave many things 
unexplained, and no hint as to where to find the reasons. 

Storing the "coirplete" document means more than showing just 
enough detail for a trained crew to interpret. Flight personnel that 
get reassigned, engineers that rotate through and work every third 
flight, new customers trying to understand how their similar operation 
might go, or managers that are troubleshooting need to be able to grasp 
the background and change history in a convenient manner. 

The attributes of a checklist that make it a precise and 
portable document for successfully operating complicated equipment also 
require a substantial effort to ensure accuracy and completeness. With 
less and less time between Shuttle flights, and the assumption that 
Space Station operations will use an entirely electronic flight data 
file, there is plenty of motivation to apply automation to the process 
of maintaining procedures. 

A Solution 

Storing the information contained in a procedure in electronic 
form for rapid distribution is the first step in automating its 
processing. A microcomputer with floppy diskettes that can be copied in 
a few minutes works for small groups, while a minicomputer tied into a 
local area network serves that purpose for a larger operation. Those 
are both recent alternatives to using a timesharing a terminal tied to a 
mainframe. A local area network, or remote computers tied by modems 
over voice grade phone lines have the advantage that not all equipment 
used has to be identical. The increased flexibilty brings with it the 
requirement that the files being transferred have a form that different 
machines can process. If the files can be built using only characters 
defined in the ASCII (American Standard Code for Information 
Interchange), so much the better. 

ASCII files can be processed by most equipment, but a simple 
text file doesn't have the structure needed for properly storing all the 
information that makes up a procedure. The trick is to have a subtle 
form of structuring the data that will let the goals be accomplished, 
but still allows quick processing. The chosen method is, of all things, 
line numbers! Or step numbers, in the case -^f a pure procedure. Not 
just plain line numbers (100, 110. infinitum) but numbers designed 


. 'i.'>' •-.■ 


to shorten the process of pagination, showing changes, making a table of 
contents, or building a list of effective pages. 

In contnist to the line number editors that have fortunately 
been replaced by screen editors, these numbers are used only Internally 
by the program. The step number can be used as a link in a technique 
similar to a relational database to show and order the steps and tie tiie 
text itself, rationale, proposed additions or deletions, all together. 
The actual steps and the associated information become a database with 
one of the output forms being masters for publishing the book. 

Figure "i • Record Structure 

■\/' -— v-^ ^Z" 


Line number ^ Text 

Each "record" in this database that makes up the procedure has 
an identical form, no matter what it represents. Of three fields, (see 
Figure 1) the first is the line number, the second a number which tells 
what type of information is contained in the text (80 characters, or 
whatever is appropriate) which makes up the third part. Such a record 
structure is simple to write out to, or read from, an ASCII file. 

One way to construct the line number is to have seven or eight 
individual numbers in array. The first few might represent the major 
sections of a book, and successive subdivisions that are used to 
organize the book. The last two numbers are needed to show the version 
number of the line, and which segment of the line is represented by the 
text in that record. 

Figure 2 shows an example line as it appears in a checklist, 
with its associated rationale. If there were more text than could be 
fit in one record, an additional record labeled as segment 2 could be 
used. Notice that the actual checklist line and the rationale both 
have the same line number, but their "types" are different. Table 1 
shows an example list of "types" that might be used for the records. 
The table can be built in whatever manner is convenient. 

Figure 3 shows the modified records for a change that has been 
requested. To show the addition, the original line is broken up into 
two p'lecec, and has the version number incremented. The text to be 
inserted is marked as being a different type, and a separate segment. 
Using some display technique the proposed insertion can be contrasted 
with the original text. An additional record is used to show the formal 





change request number, which can be tracked for that new version. If 
desired, additional records can be used to show who requested the 
change, or what it's approval status was. 

Figure 2 - An Example Step 

MSI CRT Report "pin out" times for PIVOT pins 


1 Report "pin out* times for PIVOT 

pins I 

The text for that step 

3 6 2 8 3 7 11 y MSI 

The individual who performs the action 



Tne equipment used to perform the action 

3|6|2|8|3|7|l|l| 10 I Used by the ground to calculate a 

3 I 6 I 2 I 8 I 3 ! 7 ! 1 I 2 I 10 [ predicted time foi the final pin 
Rationale for that step 

Defining type 5 as a switch name, and type 6 as its cornmanded 
position provides a means of checking nomenclature and tracking switch 
throws. When the procedure is being written, and a type 5 is input, it 
can be checked aga'nst a list of switch names, to be certain that is an 
existing one. That same list could be used to define all the possible 
positions for that switch. The structure also allows the handy ability 
to find and highlight all the callouts for a particular switch or set of 
switches. Designations for switch panels (or racks of equipment) can be 
used in the same way. By maintaining the sequent e of switch uses, it is 
possible to tell the present configuration of a panel, assuming you knew 
what it was when the procedure was sta-ted. 





Table 1 - "Types" of Records 







Plain text 


Blank line 


Inserted text 


Deleted text 


Switch name 



Switch position 
Individual performing step 
Equipment used to accomplish step 
Approval status 


ID of change request form 
Initiator of change request 


Figure 3 - Changes to the Step 

After change has been requested: 
MSI CRT Report "pin out" times for PIVOT& KEEL pift. 



Report "pin out" times for PIVOT 

3 6|2|8|3|7 2J2 

3 1 &KEEL 

3|6 2!8i3|7|2|3 



The new line with the proposed addition 


11 SYNDPY-15 

The identification number of the formal change request 


3|6|2|8|3|7|2|1 9 

Not approved 

Status of the change request 




A record of type 7 ifight be the name of the crewman that 
performs the action. The database can be quickly scanned, finding the 
particular crewman's steps (the next one or all of them) to be 
highlighted. On the occasion that different crewmen's steps are put on 
separate pages, (Commander and Pilot on the right-hand sid9, Mission 
Specialists on the left-hand side as is done i.i some books ) they can be 
easily segregated. 

With the different levels of organization embedded in the line 
numbers, some editing processes can be done uutomatical ly. Generating a 
table of contents requires only fetching the first record within each 
minor heading, or whatever level of division is desired, and listing 
those titles consecutively. As changes are made, the list of effective 
pages can be determined with similar ^ase. Pagirition rules can be 
defined that will avoid breaking up blocks of related information 
(paragraphs in plain documents or underlined procedures in a crew 
checklist) as desired. 

Lines need not appear on the page in the order " their line 
numbers. An ordering table can be maintained by the program to keep 
track of the sequence in which the steps appear. This table is also 
used when the screen cursor is moved, to relate that coord inat"" back to 
the step being touched. 

User Interface 

That logical method to tie all this infor-mdLion together is 
nothing really new, or particularly clever. The ether half of the 
solution, the hardware-dependent techniques now availdble to present the 
numerous pieces to the crew and other readers in an uncluttered manner, 
is what makes it vorth talking about. Several hundred kilobytes of 
memory, and fast processors in a package cheap enough to be distributed 
to individual offices are the ke^ to making an automated solution 
practical. Mic; ocomputers with enough memory to easily support the use 
of color and "windowing" software provide a convenient way of displaying 
the different levels of information that form ? checklist. 

Additions or deletions can easily be distinguished from 
unchanged text with the use of colors. A different background shade can 
highlight the step currently being examined. Hues can also separate the 
actual procedure being studied from the statusing inforriiation and editor 
command options. If there are a limited set of type fonts used, they 
can be indicated with a change in color. 

For this discussion, windowing is the ability to overlay a 
portion of the physical screen with a different background color and 
fill it with ancillary text. This can be done more than once if needea, 
using different shades, or lines to form a border to separate successive 
windows. When any particular window is no longer needed, it can be 
closed with the stroke of a function key, uncovering the information 
pi-eviously shown below it. Windows, in this case, do not allow the 
computer to be performing distinct tasks simultaneously. 

.-it 410 

Figure 4- Editing Display 


fl H»lp i2 Comniind* f3 Rational* f4 Chang* HlttoDr 

410 Exit 

MS> SSP ^tb KEEL PIN IN - bp <*«l*r 3 •*c> 

A4t*r approx 3 m'O 

Record Tim* i 

KEEL mtr A(B> OFF - ITD' I3(!6> *0 EXEC 

Continu* pin p*tpaction on a) t*rnat> 
motor until PIN STATUS A(,B - OUT 


1 1 

R*port -pin out" tl(r»» ♦or PfVOT 4 KEEL pint 

1 1 

R»cord pr*dict*d PUSH-OFF pin tim* on 4-6 


While working on a checklist, *-he workstation screen can use the 
layout shown in Figure 4, though it (and the next figure ) can't show 
the use of color. The main area of the screen is used for viewing 
portions o*' a checklist page, with another area needed to show function 
key use. Conmiand and "Status area can be kept to a minimum, since a 
window can easily be called up to give more detail. As the cursor 
passes over each step, that step is highlighted by a change in back- 
ground color. (For these figurps, lines around the procedure step take 
the place of contrasting colors to indicate th3 "current" step.) Though 
the computer cannot display all of a printed page on the screen at one 
time, status lines can be thrown up at the top and/or bottom on demand, 
to identify page numbers or headers that are associated with the page. 




A keystroke creates a window containing rationale or a 

provided, or a history of changes to the step, right on the 

as shown in Figure 5. The windows have overlaid the 

present, but like a sheet c^ paper covering a corner of 

can be whisKed away to reveal the original wording. All 


background linked to a particular step can be conveniently shown, as 
needed. For a particular switch name used in the step, a window might 
spnll out the full name of the function, and notes on its use, since the 
database can be queried about such things. This is the electronic 
equivalent of holf^ing your finger at one particular place on the page, 
and piling other documents on top of it. But what is the advantage of 
doing it this way? 



.^ ,-^ 

Figure S - Display with Rationale and Change History 

*\ Help *2 Coimiandt 43 Actional* i* Chang* Hittor>> 

«10 Exi t 

MS2 SSP ^tb KEEL PIN IN - bp <«<»«r 3 %»c> 

Aft»r approx 3 mtn 

Racord Tim» i 

-Rational •- 
Ut»d by tha ground to calculat* a 
prtdtcttd tint* for tht <ina\ pin 

KEEL mtr A<B> OFF - ITEM 13<1«) ♦O EXEC 

Continu* pin retraction on al ttrnat* 
motor until PIN STATUS AdB - OiJT 


-Chang* hl»tory- 
SYN OPY - 13 
••Not approv*d»» 


R»port 'pin out' tim*» for PIVOT fc KEEL pin* 


R»cord pr»dict*d PUSH-OFF pin tim* on 4-6 

When requesting a change to an existing procedure, the engineer 
can see the up-to-the-minute status of the book, including other changes 
that are pending. Once his change is made to a central copy of the 
book, all others interested in the book can see that request, 
immediately . People at remote locations can see the proposed change, 
reasoning, and comments as fast as it can be electronically transmitted. 
Changes since the last edition can be flagged as to whether or not uhey 
have been approved by the whole community, or are still under 
consideration. Since the rules for formatting are contained in the 
routine for displaying a procedure, the result will be consistent with 
standard practice. If approved, the changed page itself and the list 
of effective pages can be updated with checking being the only 
additional human work. 

When a change is proposed, other information about that request 
can be gathered for tracking by management, if needed. Reasons for the 
change, books affected, initiator's name, concur»-ences, dates, and other 
data can be recorded for use in separate programs. 

When creating a new procedure, rationale for the steps can be 
captured and preserved when the steps are entered, since it is 
convenient to do. Oatabasp-like functions, now available with an 
electronic document, can also be used when building the checklist. 
Status information about what panel, software block, or digital 
autopilot mode is being used can be shown, or added to the top of a new 
page. The total result can be distributed immediately after creation 
and approval . 



It appears that the burden for building the table of contents 
and formatting the document falls on the engineer writing the 
procedure, but that is actually a hidden part of the input process. 
Windows, if applied copiously, can be used to give clear information 
about command choices. A "help screen" can be used to explain use of 
the function keys on the workstation to perform a "block move", for 
example. In short, the environment can be made simple enough for even 
an en gineer to uce! 


The equipment needed to provide this automated tool isn't 
particularly fancy. Most business microcomputers have at least the 
native ability to handle printing letters in different colors with that 
same palette available for background colors. A rich character set that 
includes a checkmark, or the graphics characters used to provide borders 
for tables can make the process easier. Most recent microcomputers have 
a hard disk, at least as an option, and room for plugging in several 
hundred kilobytes of memory. For a miniscule cost in purchased 
software, the basic building blocks are there. 

This prototype was built using, an AT & T 5300 (an IBM compatible 
machine) equipped with a hard disk, the full 640 kilobytes of memory, 
and a color monitor. The language used was Turbo Pascal, sold by 
Borland International. The windowing software is a package called 
"Virtual Screen Interface", a product of Amber Systems, Inc. The entire 
package described has a list price near $5,000. 


Storing proced res (and other documents that combine the work of 
several people) for electronic distribution and editing greatly speeds 
the process of coordinating changes and producing the final book. The 
methods outlined can be used to show suggested changes and the rationale 
that supports individual sections of the document. Modern hardware, and 
innovative software now available provide clever ways to display the 
information, as needed. By treating a procedure as a database of 
actions, times, people, and equipment, instead of just a text file, 
convenient methods can be built to understand the operation. Embedding 
editing guidelines and procedural nomenclature checks in the program 
makes the process of creating or changing a checklist faster and easier 
for the personnel involved. 

Mr. Hollingshead is a Senior Engineer in the Orbit Procedures 
and Flight Data File section at JSC, building satellite deploy 
procedures and pr-oviding automated tools for checklist maintenance. 
His prior experience includes running and writing instructions for the 
functional, qualification, and acceptance testing of varied aerospace 








Ronald K. Petersburg, McDonnell Douglas 
Technical Services Company, Inc., Houston, Texas 



The McDonnell Douglas Corporation has a Corporate-wide program 
to create a management and work environment conducive to increased job 
satisfaction and continuous quality/productivity improvement. A re- 
lated activity called Streamlining is an integral part of the McDonneT. 
Douglas Technical Services Company-Houston Space Transportation System 
^ Engineering and Operations Support (STSEOS) Contract with NASA/JSC. 

a The paper will discuss the development of Streamlining as a 

W. concept, its inclusion in the STSEOS Contract, and how it serves as an 

^ incentive to management and technical support personnel. The mechanics 

P of encouraging and processing Streamlining suggestions will be discussed 

I including reviews, feedback to submitters, recognition, and how indivi- 

f dual employee performance evaluations are used for motivation. Several 

,i items that have been implemented will be mentioned. Information re- 

ported to JSC will be discussed as well as the methodology of determining 
estimated dollar savings. 

Finally, the overall effect of this activity on the ability of 
the McDonnell Douglas flight preparation and mission operations team to 
support a rapidly increasing flight rate without a proportional increase 
in cost will be illustrated. 


'' The Space Transportation System (STS) has provided a vehicle to 

deliver, retrieve, and service a wide variety of payloads in the near- 
earth orbital environment. Many design, fabrication, implementation 
and operational challenges have been met and undoubtedly there will be 
more to come. The NASA and its Support Contractor team have taken the 
STS from a developmental environment into an operational era. Support 
for the early flights was extremely labor-intensive since the flight 
objectives were designed to systematically verify the operation and 
\l flexibility of the vehicle systems. Starting with the fifth flight, 

-J however, emphasis shifted to operational objectives with the addition 




of paying customers and valuable commercial and scientific payloads. 
This very quickly brought home the fact that the STS must become proven 
and cost-effective to compete with other delivery systems. 

McDonnell Douglas Technical Services Company (MDTSCO), through 
the Engineering and Operations Support Contract, is playing a signifi- 
cant role in transforming STS from a R & D to an operational system. 
McDonnell Douglas provides engineering support for flight preparation 
and mission operations. Recognizing these activities ds a signifirsnt 
factor in the cost per flight equation, McDonnell Douglas program man- 
agement devised a process to provide cost-effective, higli quality 
support in a rapidly increasing flight rate environment. 

The approach is called Strsdmlining and the remainder of this 
paper will further define t'ne term, discuss how it evolved as an effec- 
tive quality/ productivity improvement tool, define the mechanics of 
tracking, reviewing, and reporting, and will explain how McDonnell 
Douglas has developed it into a never-ending process of continuous 
quality/productivity improvement. Also included will be examples of 
results and estimated savings. 


An exact definition of Streamlining is elusive and can vary 
greatly froin government agency to agency, company to company, manager 
to manager, and person to person. There is nothing wrong with this as 
long as the ultimote goal of such an activity results in increasing 
productivity without sacrificing quality. 

In broad, but appropriate terms, the McDonnell Douglas contract 
defines Streamlining as "initiative/innovation in reducing the cost of 
STS flight prepax'ation and/ or increasing effectiveness of STS flights". 

McDonnell Douglas has not limited Streamlining to direct con- 
tract support responsibilities but has provided concepts, ideas, and 
implementation plans for productivity improvements to benefit the over- 
all STS program. 


McDonnell Douglas determined that four basic ingredients were 
essential to successfully develop and implement a Streamlining process: 
1) Management commitment; 2) Providing the prope»~ environment for the 
workforce; 3) Incentives for the workforce to participate; and, 4) A 
system for recording and reporting progress (internally and to the 
customer). All of these are equally important for gaining m.anagement 
and workforce support and customer awareness and acceptance. They in- 
still a cultural shift toward high-quality products at the lowest 
possible cost. 


.yt-^Mk^:^.- :?. 

\-« ■•«. >■ 'NIC 

Manageiiient Commitment 

A lack of top-level management commitment to any new initiative 
can mean failure of that thrust. This is true whether it is only per- 
ceived by the workforce or if it is seen by the workforce as being real. 
Leading by example has been found to be a powerful tool. Continuous 
encouragement down through the chain of command serves as a constant 
reminder to the workforce. 

Providing The Proper Environment 

Engineers and scientists need more than a desk, paper, pencils, 
access to a computer, and a periodic paycheck to become enthusiastic 
streamliners. The workplace environment at McDonnell Douglas is made 
more conducive .d Streamlining by: 

1) Encouraging skill sharpening through in-house and advanced degree 

2) Applying paraprofessionals to necessary routine, repetitive, manual 

3) Implementing standardization and then automation to known processes 
and/or product generation; 

4) Providing employee involvement through the application of Partici- 
pativp Management techniques; 

5) Supporting the formation of problem-solving teams (Quality Circles) 
in work areas - along with appropriate leader training in brain- 
storming and ether problem-solving techniques; 

6) Increasing the availability of micro-computers for general engineer- 
ing use and providing adequate remote access terminals to large 
mainframe computers; 

7) Making management more visible through boss talks and workplace 

8) Reinforcing the cuncept that there is always a better way to accom- 
plish a task or to improve a process; 

9) Assuring the workforce that attention to Streamlining or eliminating 
their assignment will not jeopardize their future but rather allow 
their talents to be used for more creative work; and 

10) Briefing all new employees on the concept and process of Stream- 
lining as part of the McDonnell Douglas orientation session. 

Following are examples of McDonnell Douglas progress rr results 
during the six-month period ending in March 1985: 

1) The number of Quality Circles grew from 9 to 22; 

2) Strong emphasis on the benefits and proper use of participative 
management techniques continued through an in-house workshop given 
to all-hands; and, 

3) The use of paraprofessionals on routine/production tasks was in- 
creased from 24 to 44. 


■--i*<..t,- ... 


Incentives to Participate 

The vast majority of employees in a professional, knowledge- 
based workforce are self-motivated to excel at their assignments. At 
McDonnell Douglas, proven incentives ar? used to focus the workforce 
talents on continually improving the process. 

Streamlining suggestions that are implemented and result in tan- 
gible, and sometimes intangible, benefits are recognized through an 
employee recognition program. This recognition takes t!ie form of an 
individual, or group, achievement certificates and/or monetary awards. 
Also used for incentive are "extras" as part of the overall compensation 
package - bonus plan payout and special salary reviews. 

Other more subtle methods include modest monetary awards through 
the McDonnell Douglas suggestion program, candidacy for the JSC launch 
honoree program, recognition by various and ether publications; e.g., 
the company newsletter and local new papers. Letters or other forms of 
recognition by the customer anH/or end-user have proven to be extremely 
powerful motivators but arc outside of McDonnell Douglas control. 

As a measure of the quantity of these motivational methods at 
McDonnell Douglas, during a six-month period ending in March 1985, the 
Engineering and Operations Contract recorded the following; 

1) More than 200 coTiinendations/awards were made as part of the employee 
recognition program; 

2) Monetary awards of about $3850 for sixteen employee suggestions; and, 

3) Twenty letters received from JSC management commending the perfor- 
mance of seventy individuals. 

System for Recording and Reporting 

Another ingredient in the McDonnell Douglas Streamlining process 
is a system to record, track, and report on the disposition of sugges- 
tions. It is also necessary to formally review the items for appropri- 
ateness and applicability. It is also important to recognize employee 
participation by direct feedback - keeping the submitter aware of where 
the suggestion is in the review process flow and, if it is rejected, to 
supply the reason. 

It should be noted that there have been suggestions that re- 
quired customer acceptance and renuired changes in policy, rules and/or 
procedures before implementation. Because of this, the end-user is 
included in the process. 

The process at McDonnell Douglas can be described by a flow 
diagram, as shown by Figure 1. 






( • 






. 1 



















A micro-computer data base Is utilized to provide summary 
reports plus feedback and acknowledgement letters. It also facilitates 
regular reporting on the progress of Streamlining to management and the 

Another important step in reporting is the interface with JSC's 
Productivity Improvement Program. McDonnell Douglas has submitted 
several ideas to JSC which were subsequently implemented and reported 
on in the JSC Crier newsletter. 


The continuous improvement approach to day-to-day and flight-to- 
flight activities has generated literally hundreds of Streamlining 
suggestions/ideas. An appreciable percentage of these have been approved 
and implemented. Collectively they have been shown to accumulate into 
sizeable estimated contractor and STS Program savings. Results have 
been formally presented to the JSC Technical Manager of the McDonnell 
Douglas Contract semi-annually since October 1982. 

It should be noted that there have been implemented ideas that 
are intangible in nature; i.e., cannot be translated into programmatic 
dollar savings. Some of these have enriched the quality of worklife, 
which certainly contributes to employee morale and motivation - and 


•■*' " "^SSP 



thereby, hopefully, lowering attrition, or turnover. Others have re- 
sulted in higher quality products by using procedures/approaches that 
are designed to significantly reduce the chance for human error. 

In an environment of a rapidly increasing 
Douglas has concentrated a great amount of effort 
flight preparation and mission operations support 
of some of the implemented Streamlining ideas are 
mainder of this section. 

flight rate, McDonnell 
on Streamlining the 
functions. Results 
discussed in the re- 

Flight Preparation Results 

Data on the McDonnell Douglas - Mission Planning and Analysis 
Division/JSC flight-to-flight reconfiguration efforts (labor hours and 
large mainframe computer time) are gathered and monitored to measure 
the effectiveness of the Streamlining results. Implemented results 
such as standardized (seasonal) Initialization-Loads, the application 
of the direct insertion ascent phase, and simplifying the analytical 
payload integration functions have contributed to the positive trend 
of flight-to-flight reconfiguration support requirements. Figures 2 
and 3 depict the results of such efforts logged for eleven (11) STS 
flights, starting with STS-5, for labor hours and computer time used, 


250 p 







5 6 7 8 9 illB lie m '♦IF 11G 51A 




l%rtMI»Tft»««*.'--^-. -i' '■-■•'•v 



9 '!iB hIC 010 tlF DIG 5W 



This information clearly shows the effectiveness of streamlining 
the flight preparation process. The data for flights SrS-5, -9, and 
STS-41C are considered "outliers" in that the objectives and/or complex- 
ity of the missions demanded considerably more effort in the preparation 
activities. They were "first-of-a-kind" flights - first direct inser- 
tion ascent, first rendezvous, and first high inclination. Some author- 
ities assume both the labor hours and computer time required per flight 
Will level off. This suggests that an increasing emphasis on Stream- 
lining would be a diminishing return situation. 

Another dramatic illustration of an increase in productivity is 
shown in Figures 4 and 5. These figures display the number of missions 
in the preparation cycle, the total labor hours expended, and computer 
time used versus time. 



g 13 


• ■^^.^ 

I 1 ' 2 ' 5 


120 i 


to 5 

198J 198<( m'' 




>■!>?»> I.*' 





utntriGiMAiiofi comiTtK usage Hi^imr 

-*- ,r' 

r AVAlLABlt^ 




"• rr T ' 5 % I r ? ' 5 ' ■< I 1 ' 2 "' 3 ' 1 I 1 ' ^ ' 
14S2 1983 198<< 1985 


As illustrated, while the number of missions in work has in- 
creased significantly, the level of labor required seems to be stabil- 
izing. Computer time requirements, however, are not so well-behaved. 
It may be that this should be expected and might be indicative of such 
things as increased automation, mission-compiexi ty, schedule slips, 
and remanifests. 

Mission Operations Results 

The McDonnell Douglas involvement in the development of flight 
procedures, operational products, crew and support personnel training, 
simulation support, and real-time mission control has been intensively 
analyzed for potential Streamlining ideas. 

One recent accomplishment that resulted from a Streamlining 
suggestion was the development of a standardized training plan. This 
satisfies the need for a structured method for developing ground sup- 
port personnel into technically competent and certified instructors, 
developers, flight controllers, or flight directors. 

The use of existing software tools for the automated documenta- 
tion of Shuttle Missi:^-" Simulator applications software was anoth r 
example of increasiny productivity and greatly reducing the chance for 

McDonnell Douglas standardized and simplified the Crew Activity 
Plan. This has allowed proper support of an increasing flight rate 
without any significant increase in staff. 


''Z'^SS^ '•''' 




Increased automation of the Flight Data File (FDF) product 
generation has eliminated much of the manual efforts associated with 
this activity. This also allows quick access to the FDF for reports, 
schedules, and updates. 

These are only a few of the Implemented Stream'iining suggestions, 
in the mission operations support area. The Mcnonnell Douglas emphasis 
on productivity improvements will ensure a continuing stream of new 
ideas and suggestions. 


f. Formal presentations are given by McDonnell Douglas to the JSC 

,| Technical Manager on a semi-annual basis. These presentations review 

» the entire Streamlining activity for a six-month period, highlight the 

most significantv and provides an estinate of program savings resulting 

from implemented suggestions. 

The process of converting labor hours, computer time, and other 
expenditures to program savings in djllars per ypjr is difficult and 
the accuracy can certairly be challenged. However, a costing criteria 
f has been developed and is consistently applied. Thr reported savings 

C are felt to be representative. The projected yearly cost sa*^ings are 

determined using a yearly launch rate of 12, two of wt.irh are Spacelab 
and two involve a rendezvous. 

The implemented streamlining items reported for the period from 
April, 1984, through March, 1985 res^ult 1n an estimated future yearly 
cost savings of approximately $8,5 K (in 1985 dollars). 


Streamlining has proven to be an effective tool to contin-jously 
improve the quality and decrease the cost of performing engineering 
support services activities like those currently underway at McDonnell 

To be successful, however, several basic ingredients are re- 
quired. There must be a top-down and demonstrated management confiiitmant, 
an environment for the workforce which is supportive of the commitinent, 
incentives to encourare workforce participation, and a S/Stem for the 
formal gathering, reviewing, and reporting of progress and achievements. 
The absence, whether perceived or real, of any of these will jeopardize 
the entire process. 

It has been demonstrated by McDonnell Douglas after a relatively 
short period of time that this approach to increasing productivity can 
develop a work environment in w!>ich continuous improvement and holding 
the gains is a way of life. 


Ronald K. Petersburg 

Program Manager, Engineering and Operations Support, McDonnell 
Douglas Technical Services Company, Inc., Houston Division, 16055 Space 
Center Blvd., Houston, lexas 77062. 

A graduate of Iowa State University with a B.S. i.- Aerospace 
Engineering, he has over twenty-five years of varied industry experi- 
ence. Sixteen years has been in direct Contractor support of the U.S. 
Manned Space Program in many engineering, supervisory, and management 

■4 ^24 




i l^ii»»;Wv^!tfk.V 

■>>* . Mn»f\ ' 




G»?rald E. Miller 
McDonnell Douqlas Technical Services Company, Inc. 
Hou'.on Astronautics Division 


To improve the productivity of astronaut crew 
instructors in the Space Shuttle program and beyond, the 
following actions are proposed. Instructor certification 
plans should be established to shorten the time required 
for trainers to develop their skills as well as improve 
their ability to convey those skills. Members of the 
training cadre should be thoroughly cross trained in their 
task. This provides better understanding of the overall 
task and greater flexibility in instructor utilization. 
Improved facility access will give instructors the benefit 
of practical application experience. Former crews should 
be integrated into the training of upcoming crews to 
bridge some of the gap between simulated conditions and 
the real world. The information contained in lengthy and 
complex training manuals can be presented more clearly 
and efficiently as computer lessons. The illustration, 
animation and interactive capabilities of the computer 
combine for a most effective means of explanation. 


As the Space Shuttle becomes an operational sys- 
stem, eveiT increasing numbers of demands of continually 
greater complexity are being placed on all areas of sup- 
port. This is easily evidenced by the ambitious launch 
schedule for the immediate future and the mission scenar- 
ios of the immediate past. The current launch manifest 
calls for a Shuttle mission each month. Assuming an aver- 
age mission length of five to seven days, it may be seen 
that twenty-five percent of support personnel's time will 






be devoted to real time on-orbit support. While simultan- 
eously, the same personnel are tasked with readying the 
upcoming crew for the next month's launch. Looking back up- 
on previous missions reveals even greater demands. The un- 
foreseen contingencies involving the Westar, Palapa and 
most recently, the Leasat satellites, demonstrated the re- 
ality of unplanned support requirements. Few tasks 
are as directly affected by these realities as that of as- 
tronaut crew training. 

Instructor certification plans 

Education in che field of hardware/human inter- 
actions, which represents the apex of crew training, will 
require constant productivity improvement if it is to keep 
pace with the operational world. This translates into bet- 
ter application of existing resources. Of course, the pri- 
mary resource of any instructor is human. One means of im- 
proving the productivity of this commodity is the estab- 
lishment of, and adherence to, a definite and structured 
instructor certification plan. 

When new employees are hired to become trainers, 
it is not unusual for them to lack firsthand knowledge of 
the system on which they are eventually expected to become 
an expert. This is especially true of the more exclusive 
systems found in the space business, such as Extravehic- 
ular Mobility Units (EMU's), Manned Maneuvering Units 
(MMU's), or evei. the Shuttle itself. Incoming instructors 
cannot be expected to have prev/ious experience with sys- 
tems such as these, since the systems are unique to a li- 
mited section of the industry. A similar problem exists for 
any area of the industry which deals with relatively un- 
common systems. As a result, newcomers often spend a great 
deal of time and energy in the familiar, "coming up to 
speed", process. The reasons for this are often quite 
common. There are always the technical manuals to be read. 
Such documents are frequently written in a fashion that is 
tedious to read and usually are not kept current with 
changing data. There is also very seldom an organized or- 
der in which manuals should be approached. There are us- 
ually training sessions or classes which may be attended 
as an observer. Unfortunately, new hires are often inun- 


• ll'^f^Sr; 


dated by the amount of information covered by an actual 
training session and are unable to interrupt with ques- 
tions since the sessions are intended for the crew being 
trained, not for the observers present. Complicate these 
problems with the fact that a single source is often not 
available with which newcomers may gauge their progress, 
and tne impedances to productivity become clear - 


A solution to these difficulties can be the de- 
velopment of a definitive instructor certification plan. 
The concept is quite similar to the curriculum outlines and 
course catalogs used by universities. All training, whet- 
her in the form of manuals or live sessions, should be 
organized into a coherent, chronological flow. A brief 
description of what is to be covered and how long is us- 
ually required should be included. Particular milestones 
should be established throughout the plan to mark times 
when an individual achieves designated plateaus in the 
field. An example of such plateaus for instructors is that 
time when a person is deemed art expert in a particular 
facet of the subject being studied. For instance, in the 
field of extravehicular activity (EVA), a "Kennedy Space 
Center Expert" designation is made when all training con- 
cerned with the launch facility is completed. 

As an example, figures 1 and 2 show the curriculum 
list and course descriptions for the Kennedy Space Cente- 
section of the EVA/Crew Systems Instructor Training Flow 
and Certification Plan. 


(Flow Code: T KSC EVA/PL) 


KSC EVA 2101- 




OJT [hST. 

OG-101-«C 1103 

QG-IOZ-KSC 1103 

QG-12S-KSC 1103 

QG-lZ6-KSr 1103 

qG.128-KSC 1103 

QG-ISO-KSC 1103 

KSC EVA WD 2131 

0-EVA EVAL 2131 

10 ! 

C-KSC EVA 2101 



'•--'#*«&« -. '"' 








TITLE: KSC EVA Monitor Duties 

DUR.ATION: 1 hour 

SYNOPSIS: This lesson Is d briefing on the duties and responsibilities of 
the KSC EVA Monitor to Include an overview of the walkdown at 
KSC, procedures for coordinating vehicle transportation/ 
logistical support, range safety considerations and scheduling 
access into the Orbiter, and budging/access training. 

Z. STPC COOE: QG-101-KSC 1103 

TITLE: Fire Protectloi Safety Orientation 


SYNOPSIS: This class fainiliarlzes students in operational areas at 

Kennedy Space Cener with fire suppression equipment, fixed and 
portable systems, and the most effective use on different 
classes of fire. 

MOTE: This class is required for KSC area access badging. 

3. STPC COOE: QG-102-KSC 1103 ' ' 

TITLE: Toxic PropeMant Safety 

DURATION: 1 hour 

SYNOPSIS: This class provides a general knowledge of the toxic propel- 

lants used at KSC. It covers the nature of the propel lants am* 
*he hazards involved in the use of the propellants. Access 
, jntrol procedures, wjirntnq systems and safety equipinent are 
covered. It familiarizes persons with the location, use and 
limitations of tm rocket propellant gas mask and air cap<;ule. 

NOTE: This class ij required for KSC area access badging. The 
refresher class (QG-106-KSC) is required ever) 3 years. 

STPC COOE: 0C-125-KSC 1103 

TITLE: Launch Comp'ex 39 Facility Safety Faraillarizatior 

DURATION: 0.7 hours 

SYNOPSIS: This class provides an overview of the facilities at Launch 

Complex 39 (LC-39) and the work flow of the Orbiter from touch- 
down to 'aunch. It covers the safety hazards and hazardous 
areas at the Shi^ttlc Landing Facility (SLF), Orbiter Processing 
Facility (OPF), and the Vehicle Asse'nbly Building (VA8). It 
also covers evacuat*— signals, egress routes and safety 

NOTE; This class is required for KSC area access badging. 


Upon completion of all necessary requirements, 
certification as a qualified instructor may be granted. 
This helps establish self-confidence within the new in- 
structors as well as the confidence of those they will be 
training. The definition of set criteria for instructor 
certification not only reduces the time required for 
trainers to develop their skills, but vastly improves the 
ability to relay those skills. 





' s- 

* ■ 


Cross training 

After an instructor is certified in one area of ex- 
pertise, the next means of productivity improvement may 
be applied. That is the concept of cross training. While 
technical proficiency is best maintained by focusing on 
a particular field, for instance extravehicular activity, 
instructors should not be disadvantaged by isoxation to a 
particular subset of that field. The impact of this ap- 
proach on training effectiveness is to greatly reduce the 
integrated flow of a task. However, by cross training 
staff members in the various subsets of their field, a number 
of productivity improvements will be seen. Among them is 
the continuity of beginning to end contact by a crewmem- 
ber with a single trainer. 

When a crewmember is being trained for a particu- 
lar task, the training is usually separated into certain 
components. This a common approach to solving any complex 
problem. The individual components, however, commonly are 
the responsibility of different instructors. The problem of 
bringing the separate information together is frequently 
left to the crewmember. Questions concerning the relation- 
ship of events in different components must be resolved 
through the interface between segregated information 
sources. Productivity is thus hampered when instructors are 
unfamiliar with the workings of other sectibns of their 
task. If instructors were adept in all aspects of their task, 
which is the expected result for the crews they are train- 
ing, then continuity could be provided from initial 
training to real time execution. This can provide crewmem- 
bers with a single, au^horitative source of information on 
their expected performance. Also, the instructors' under- 
standing of the overall task will be greatly improved. Due 
to this improved understanding, the quality of information 
passed on by the instructor will increase. This is often man- 
ifested as a keener insight into the complicated relation- 
ships between the mechanical systems being used. Thus, 
productivity may be improved with respect to the time need- 
ed to resolve problem.s as weJl as the extent to which prob- 
lems may be solved. 

Further productivity improvement as a result of 
cross training may be seen in the area of scheduling. As 
Space Shuttle missions are slated with ever increasing 
frequency, the scheifuling of instructors assigned to a part- 
icular mission becomes continuously more delicate. Add to 



^>-#Y'^.^'^ "^- " " ^ 

this the demands of unforeseen contingencies, such as salvag- 
ing the Westar, Palapa, and Leasat satellites, and the resul- 
tant loads placed on training staffs can become overwhelm- 
ing. By creating a training cadre comprised of members who 
are equally knowledgeable in all aspecLs of their task, 
the negative impact of such demands can at least be mini- 
mized. CJnex >cted work requirements that shift personnel 
away from scheduled training, as when instructors are need- 
ed to develop salvage procedures, can leave crews without 
instructors. With a fully cross trained staff, loss of pro- 
ductivity can be minimal since other instructors would be cap- 
able of taking over for those who were called away. 

Improved facility access 

With a strong understanding of the system in hand, 
an instructor's productivity may be further enhanced through 
improved facility access. This involves making instructors 
more familiar with the training environment by exposing 
them to it, not only as instructors, but as students as 
well. Instructors' ability to convey learning expectations 
is vastly iuproved if they have experienced the training 
session for themselves. 

Crew training can, in general, be divided into two 
types, that which is mentally rigorous and that which is 
physically rigorous. Often it is some combination of the 
two, as is acutely illustrated by extravehicular activity. 
In either case, it is inherently necessary to place the 
crewmenber in environments which are initially unfamiliar, 
complpx, harsh or even potentially hazardous. In most in- 
stances, instructors will have studied and observed the fac- 
ility environmsnt. But rarely will they be given the oppor- 
tunity to experience the demands it will make on the pro- 
cedures beirg developed or relayed. Instructors are conse- 
quently forced to plan tasks using conjecture, assumption 
or speculation that the procedures will be workable. Pro- 
ductivity suffers because cewmembers must often work out 
techniques from scratch. The instructors are unable to short- 
en the time required to do this since they lack the e^^per- 
ience of practical application. The overall effect can be 
compared to taking piano lessons from a teacher who under- 
stands music theory, but has never actually played a piano. 

To illustrate this point, the figure below shows 
an underwater training session at he Weightless Environ- 
ment Training Facility (WETF). All of the procedures be- 



.wi>..i**irr.^ x^ - 



ing learned by the crewraembers were developed and convey- 
ed by a trainer. Without exposure to this environment, the 
trainer must speculate about all physical aspects effect- 
ing the task. These may include overall mobility, strength, 
duration, visibility, manual dexterity within a pressur- 
ized suit or many others. 





.i I ■ ! ' 'J* \ 



It is possible to remedy this sitfjation by expos- 
ing instructors to the conditions of training facilities 
whenever feasible. While new procedures are still in the 
early developmental stages, trainers should undergo in- 
structional sessions identical to those later intended for 
use on the crews. Thus, once instructors begin presenting 
these new procedures as facts, they would command a much 
better perspective of how to approach the task. Communi- 
cation between the crewmember and instructor would improve 
since each would possess a common reference of experience. 
Productivity increases as instructors become better able to 
convey expectations, and more importantly, techniques to 

integration of former crews 

Another means of productivity improvement is the 
integration of former crews into the training of upcoming 
crews. This integration may seldom be in the form of an 
actual instructor. Instead, the role often becomes more that 
of a consultant. 

Techniques fabricated on Earth, yet intended for 
space, are frequently hindered by the simple fact that they 
were created under the constiaints of planetary limita- 
tions. Regardless of the painstaking efforts made to sim- 
ulate the space environment, many vast discrepencies still 
exist. As a result of these discrepencies, procedures which 
work well in simulated space may not in actual orbit. Often, 
the reverse is true. Tasks which were foreseen as being 
quite laborious, may turn out to be very simple in space. 
The constant use of feedback from returning crews thus be- 
comes an excellent means of improving productivity. It 
has been common practice over the years for all support 
personnel to make use of the debrief information from for- 
mer crews. It should be noted, therefore, that what is be- 
ing suggested is intended to go beyond the passive relay 
of data to a more active involvement in training sessions. 
For example, having an experienced crewmember in atten- 
dance during a WETF task often provides instructors with im- 
mediate feedback for questions concerning the manual man- 
ipulation of satellites, on-orbit crewmember work re- 
straints or maneu'.'erability about the payload bay in zero 
gravity. This accurate insight as to how well the train- 
ing simulates the real world is invaluable. 

Computerized lessons 


^X - .-■ '■fivlf^: 

ORIGINAL r ,-" : v^ 

The use of illustrated/animated computer lessons 
to supplement printed texts can produce an outstanding 
improvement in productivity. Complex systems are made much 
clearer when their function can be animated. The immediate 
feedback and individual interaction of a computer also re- 
present improvements over printed workbook/lecture combin- 

■ f. 

Nearly all Shuttle systems which crewb must learn, 
indeed all space systems in general, are to say the least, 
complicated. It is also true that regardless of their com- 
plexities, the systems must be mastered on schedule. 
Therefore, training programs are inevitably faced with the 
problem of how to convey a working knowledge of complex 
systems w:iile maint-aining the highest level of productiv- 
ity. Traditionally, the initial stages of crewmembers' 
training consists of many hours spent studying the volu- 
minous stack of technical manuals associated with their 
given assignments. Such documents are always quite thor- 
ough, yet their effectiveness can be greatly improved 
by supplementing them with computer lessons. For ex- 
ample, figure 4 below shows a schematic diagram of the 
Extravehicular Mobility Unit (EMU) used by Space Shuttle 
EVA crews. This space suit diagram is representative of 
those used in most training manuals. 


y i. l.H I . I«l. l « l.HI.IM. I .I.I.I . ,.l ,l. ., | j.j|. . ,.l.l..«...... »..|»|.1.I 1 I1H | UI...H|,|, | |. | , |,| „ . 








Normally, a schematic such as this will appear in 
a manual along with detailed drawings to illustrate each of 
its components. Down to the smallest item, the system is 
methodically chronicled by diagiams such as the one below. 
Figure 5 is an expanded view of the pressure sensor called 
out in figure 4. 


Pivot mech2nis.T^ 

Outsr case — 

Ball seal for Hems 132 & 138 
Porous filter lor item 114 

OJ4 ^ Resistive element 
(wire wound) 


-Reference pressure 
(vacuum or ambient) 

Sense port 

It may be s&3n that figure 5 illustrates one of 
the least complex components, representing the smallest 
fraction of the e.itire system. Yet, a substantial amount 
of written text would accompany figure 5 in a training 
manual to explain the sensor's interaction with the over- 
all system. lixpanding this example to the trul\ compli- 

cated components and then adding the require- 
vey their physical interaction during oper 
integrated system, yieJds an understanding 
being addressed. Not only is ^he printed te 
do this cumbersome, it is a^^.o tedious, nonj. 
and di/.ficuit to update as changes occur. 

to oon- 
i£ t-he 
■ ;! demands 
eedf -o 

Fully illustrate^ and animated explan.i of 
systems such as the EMU can be programmed into uo.nputer 
lessons. Animated representations of flow paths, fans, 
pumps, motors and electrical circuits in operation can 
convey their physical interaction faster ard more clearly 
than printed text and static diagrams. Productivity im- 
proves further since crewmembers may work at whatever 



lite**^.-'--* A>.««" 


pace suits their degree of familiarity with 'che system. 
This is done by establishing level within the pi-ogrammed 
lesson. With such an interactive pLoqram, the crewmember 
may request further explanation of particular areas, or 
move through a brief overview of well known ones. Also, 
since most complicated texts require the scheduling of a 
live lecture session to explain them, time may be used 
more efficiently. This is a result of the fact that most 
well programmed computer lessons require only a brief dis- 
cussion with d trainer to clarify points left after the 
lesson, as opposed to a lengthy lecture. Further improve- 
ment in productivi»-v w ; "" 1 be seen as system updates oc- 
cur. When changes occur . ich effect printed texts, page 
change notices must be generated and distributed. These 
must not only be manually added to all existing copies 
of the text, bui hand carried to wherever those copies 
may be. However, a single change in the program of a com- 
puter lesson can complete a" update for all who wish to 
use it, regc.rdless of thei. location. 


Productivity improvement in the field of crew 
training becomes imperative as the Space Shuttle enters 
its operational phase. Instructor certification plans, cross 
traininc. facility access, integration of former crews, 
and computerized lessons represent some of the means a- 
vailable to incease the effectiveness of instructors. All 
of the measures discussed have been adopted by the EVA 
Astronaut Training Office. The value of such measures may 
be seen in the following statistics. From 1981 to 1985, the 
EVA training office has experienced a 25% increase in per- 
sonnel. During the same time period, the n-imber of Shuttle 
flights oeing successfully supported annually has increased 
600%. It should be noted that all of the actions discussed 
are applicable to productivity improveme- t of any form of 
training, not only astronaut crews. 



\ S!!^ 




The author received e Bachelor of Science degree 
in Aeronautical and Astronautical Engineering from the 
University of Illinois incJanuary, 1983. Assignment to 
the Astronaut Crew Training Office was accepted after 
serving in the software development branch c:. the Space 
Vehicle Dynamic Sino'lation Program. The author is a cer- 
tified crew trainer specializing in the field of extra- 
vehicular activity, particularly as pertains to the oper- 
ation of Extr 3hicular Mobility Units. 


-^i'^J3^'^\- . 




■ N86-15196 






C. E. Bryan, McDonnell Douglas Astronautics Company 

D. A. Maclean, McDonnell Douglas Astronautics Company 


This paper describes how the McDonnell Douglas PAM ground 
processing operations have evolved since they were started at KSC in 
1982. The objective of the changes was to reduce the prelaunch testing 
of the Airborne Support Equipment in order to Increase the throughput of 
PAM systems while not compromising the reliability of the system when 
functioned on-orbit. The changes that resulted from lessons learned and 
experience gained from the initial cargo element ground processing, the 
on-orbit performance of the systems, plus the post-flight refurbishment 
and recertifi cation of the Airborne Support Equipment have resulted in 
significant reductions in labor expenditures and work shifts required to 
prepare a PAM system for flight. Our streamlining efforts are continu- 
ing and are expected to yield additional productivity improvements in 
the future. 


Recognition of the need for an economical upper stage to augment 
the Space Transportation System's large and heavy payload capability led 
to the development of the Payload Assist Module (PAM) as a spinning 
solid upper stage. The stage was sized to provide the capability for 
boosting a Delta Expendable Launch Vehicle class satellite to 
geosynchronous transfer orbit after release from the Orbiter in a low 
circular orbit. 

The development of a spin stabilized system represented an 
orderly adaptation of a family of spinning upper stages which McDonnell 
Douglas had designed for NASA as the third stage of the popular and 
well -proven Delta Expendable Launch Vehicle (ELV). The National 
Aeronautics and Space Administration agreed in 1977 to let McDonnell 
Dojglas develop the PAM system on a commercial basis and offer it to 
satellite system owners for payloads of the Delta ELV class. The 
proposed system had the additional advantage of being adaptable to 
either the STS (PAM-D) or the Delta ELV (Delta PAM), thereby providing 
the '■atellite owners with a backup launch capability as the STS was 
proceeding through the final phase of its development program and the 
first four STS development flights. The first of 12 Delta PAM launches 
took place in November 1980. This was followed by the launch of the 
first two PAM-D systems on STS-5 on November 11, 1982. A growth version 
of the PAM-D system was given a go-ahead in 1983. The growth version is 
designated PAM-DII, and it is scheduled to make the inaugural flight in 
November 1985. The launch record for the first 14 units along with 
those cur-Tiitly scheduled through mid-year 1986 are provided in Table 1. 

^ c: - 6 



















































STS-51 I 











n "27-85 



. TBD 








STS-61 H 










orig!a/;". . : •- 


The PAM Is a system designed to provide the necessary Injection 
velocity to deliver pay loads (spacecraft) from the low earth orbit of 
the STS Orblter Into higher energy transfer orbits. The PAM flight 
system Includes the consumable expendable vehicle (EV) and the reusable 
Airborne Support Equipment (ASE) that Interfaces with the Orblter. 

The PAM expendable vehicle hardware consists of the solid rocket 
motor (SRM) and the payload attach fitting (PAF). The ASE Includes the 
cradle, the splntable, the thermal control system, and the avionics 
equipments required to functionally Interface with the Orblter and 
flight crew. The PAM flight system hardware Is shown In an exploded 
view In Figure 1. 



Reaction Fittings 




& Separation 



Large Sunshield 



Standard Sunshield 



*i*»«N^(*^^- , V.-, - 


PAM Expendable Vehicle 

The SRM Is the propulsive element of the PAM vehicle. It is 
supplied by Morton Thiokol, Inc. and is designed to accommodate a wide 
range of mission performance requirements. Propellant loading for the 
baseline PAM-D mission is approximately 4400 pounds. The new PAM-DII 
SRK is also supplied by Morton Thiokol and has a maximum propellant load 
of 7155 pounds. The propellant lOvl in both motors can be varied as a 
mission option. 

The PAF provides the means of attaching the spacecraft to the 
SRM and the mounting accommodations for the PAM avionics subsystem 
boxes. The PAF structure has two reaction fittings to provide load- 
carrying paths to the cradle forward restraint arms. These restraint 
arms are retracted to permit spin-up of the PAM and the spacecraft prior 
to deployment from the Orbiter. The components for a fully redundant 
electronic sequencing system (shown schematically in Figure 2) are 
mounted on the payload attach fitting. The avionics system consists of 
a timer assembly, an electronic control assembly, and batteries. The 
avionics system commands SRM ignition, spacecraft separation, and 
yo-weii>rtt release. 




. I 

PAM Vehicle 



External Power 

PAM System 

PAM System Moniinrs 





Safe and Arm Poslllon Commands 

Safe and Arm Position Monitors 


Sate and 
Arm Device 

) Spacecraft 





J System 
"^ Ordnance 

^^ J 



PAM Airborne Support Equipment 

The PAM cradle is the aluminum structure which mounts the system 
in the cargo bay. The cradle is attached to the Orbiter with four 
longeron and one keel attachment points. The spintable is bolted to the 
cradle and it serves as the pedestal for mounting the motor to the ASE. 
There are two motor driven restraint arms on the cradle which provide 
the forward structural attachments to the Expendable Vehicle Payload 
Attach Fitting. The cradle also provides mounting accoiwnodations for 
the sunshield and the avionics equipment. 



The spintable provides the rotational velocity to the PAM/space- 
craft combination for stabilization after deployment. Spinning is 
accomplished by redundant electric drive motors, which are capable of 
providing a S5»lected spin rate in the range of 40-100 rpm for PAM-D or 
35-75 rpm for PAM-DII. The PAM expendable vehicle is joined to the 
spintable by a vee-block clampband. When the clampband is released by 
the redundant bolt cutters the separation springs on the spintable 
provide the separation impulse to deploy the expendable vehicle and 
satellite from the cradle in the Orbiter. 

The ASE avionics system, which is comnon to both PAM-D and 
PAM-DII, is shown in a functional layout In Figure 3. This system 
Interfaces with the STS Orbiter, the PAM expendable vehicle, and the 
spacecraft. The PAM avionics accepts and implements commands from the 
Orbiter General Purpose Computer (GPC) data bus and Standard Switch 
Panel, distributes Orbiter 28 VDC power to the PAM and spacecraft, 
provides closed-loop sequencing of PAM systems, and generates system 
status information for display to the crew and downlisted for real time 
monitoring in the Mission Control Center at the Johnson Space Center. 
The microcomputer-based Sequence Control Assemblies are the heart of the 
avionics control system. The sequence control assemblies tSCAs) are 
redundant controllers for the sequencing and operation of all PAM 
equipment in the Orbiter. Certain key steps in the sequence are 
performed by the Flight Crew (switch panel or General Purpose Computer) 
due to their safety or timing critically; however, the majority of the 
detailed sequential steps are handled automatically by the 
The SCA, with the signal conditioning unit (SCU) and the 
distribution box (SSDB), controls the operation of the 
elements, the opening and closing of the sunshield, and the 

the restraint arms, spin system, and deployment 
rates are prevented through the use of dual 


SCA system. 

spin system 

PAM heater 

operation of 

Excess spin 

overspeed protection 


Pnmf PL 



~l ' Umhhcll 

"j [~ PiyllMd "L 

'OptiOnfl PDI Port 



, v""- '^if^Ci:. 




ASE Thermal Control System (TCS) 

The Thermal Control System consists of multl layered thermal 
blankets mounted on the cradle to provide thermal protection for the PAM 
EV and ASE eq. ments. The sunshield is mounted on top of the cradle to 
provide thermal protection for the spacecraft to control the on-orbit 
solar input to and heat loss from the spacecraft when the Orbiter bay 
doors are open. Thennostatically-controlled radiant heaters are 
installed on the internal surface of the cradle blankets to provide the 
required heat input for cold STS attitude orientations. The portion of 
the sunshield covering the top of the spacecraft is a clamshell 
structure th?.t is opened and closed by a redundant electric rotary 
actuator operating a wire rope cable system. 


The PAM system is launched with the sunshield clamshells open 

and in an electrically passive mode. After the Orbiter has achieved the 

desired 160 to 190 nautical mile circular orbit, the ca' ;o bay doors are 

opened, the PAM system is powered up by the crew, and the sunshield is 

closed. After closing, the system is powered down until 60 minutes 

prior to the nominal deployment time, The Orbiter maneuvers to achieve 

the proper deployment attitude. The attitude maneuvers are completed by 

deployment minus 40 minutes. The deployment sequence is initiated by 

the flight crew at deployment minus 15 minutes. The sunshield is 

opened, the restraint arm withdrawn, and the spin system brings the 

expendable vehicle and spacecraft up to the pre-programmed spin rate. 

; At deployment minus 3 minutes, the terminal sequence is initiated by a 

f General Purpose Computer command to the SCA. During the terminal 

I sequence, the ordnance systems are armed, the spacecraft is configured 

I for deployment, and the final deployment command is issued by the 6PC. 

I The EV and spacecraft are ejected from the cradle at approximately 

■ 2.5 feet per second at the pre-programmed spin rate. When the EV is 

released, the timing systems on the payload attach fitting start a 

countdown to initiate the solid rocket motor 45 minutes later. After 

the deployment is completed, the flight crew initiates the closure of 

the sunshield to protect the remaining equipment, and powers down the 

PAM ASE. The PAM ASE is powered up again in preparation for payload bay 

door close for descent. After powering the system, the sunshield is 

opened and the system is then secured for descent. Total operating time 

of the avionics system is usually 60 to 75 minutes on-orbit. Total 

spinning time is just over 15 minutes. 


Preparation of PAM systems to meet NASA manifested mission 
schedules is influenced by Ground Support Equipment and checkout 
facility limitations and the four sets of MDC-provided PAM-D Airborne 
Support Equipment and three sets of PAM-DII ASE. At the initial 
inception of the launch site testing in May 1982, NASA allocated one of 
the two checkout bays of the Explosive Safe Area 60 (ESA 60) Facility on 
\ the Cape Canaveral Air Force Station for the assembly and testing of the 

PAM systems. MDC provided a single ground checkout test set. This test 

' 442 




set was moved to ESA 60 after being used in the factory for over two 
years of system integration design verification testing at Huntington 
Beach, California. The test set, officially identified as the PAM Model 
500 Test Set, is a mini -computer controlled "Orbiter simulator" and 
telemetry ground station. The Orbiter simulation portion of the test 
set provides high fidelity Interface simulation of the Orbiter avionics 
systems shown in Figure 3. The test set also provides test equipments 
for monitoring the avionics equipments packaged on the expendable 
vehicle Payload Attach Fitting. 

This multi-purpose test set limits testing to a single test 
article. The cabling requirements to connect a cradle or an expendable 
vehicle to the test set make it prohibitive to redirect the short 
duration testing from one cradle to another or from one expendable 
vehicle to an alternate. The test set also restricts the testing to 
either a cradle or an expendable vehicle, not both in parallel. These 
restrictions, therefore, make serial scheduling necessary. The ESA 60 
hazardous processing facility introduced several operational compro- 
mises. Including limited floor space, which had an adverse effect on the 
overall operation. Most of these compromises were eliminated when the 
PAM operations were moved to the Astrotech TICO facility in Titusville 
in the Fall of 1984. This facility was sized to accommodate the 
physical needs of the PAM equipment. . It offers a separate control room 
for housing the test set, an overhead hoist adequately sized to allow 
lifting the entire cargo element with a single hoist, more floor space 
in the checkout cell, and a dynamic balancing machine in a room adjacent 
to the test cell. These facility improvements and the inclusion of the 
dynamic balancing operation to the sa.ije building provided an opportunity 
to consolidate our operations in one location. Figure 4 depicts the 
flow of PAM hardware at the Eastern Launch Site. 




• WACfCnAf'T 





• souD nOG>9.r manm 

-^ t •mCECAAFT MU*-0 <MVUMI)) MATt 










'.-^-'^v%c:. ■ '■ 


.■i ^'3.^^ 



The Astro tech facility is now the central hub for McDonnell 
Douglas PAM ground processing activities. The PAM elements are staged 
through this facility as shown in Figure 5. Once the PAM system is 
completed, the spacecraft is added and the total cargo element is 
readied for transport to the Kennedy Space Center, Vertical Processing 
Facility. Upon receipt of the cargo element at the Vertical Processing 
Facility, NASA and its supporting contractors place the cargo element 
into the vertical test cell and verify the cargo element interfaces. 
They then move the total p^yload complement to the launch pad, install 
it in the Orbiter, certify the interfaces bet»i»een the Orbittr and 
payload, and launch it. At the completion of the mission, the Orbiter 
and the payload ASE are returned to the Eastern Launch Site where the 
cargo residuals are returned to the appropriate payload organizations. 
In PAM's case, the ASE is off-loaded to a cradle support stand, covtred 
and returned to the Astrotech facility for post-landing inspections, 
refurbishment, reconfiguration, and recertlfication before it is 
reloaded with another expendable vehicle for the next usage. PAM Unit 
11, prepared for the ANIK-Cl spacecraft, was successfully processed 
through Astrotech in November and December 1984. This system 
launched and deployed from Discovery in April 1985. 




















B ^A\. 

























,---- 'V*,*-^.- 


>^, .*^kkll-.v 


• ^ 



After processing the first two units for STS-5, an effort was 

^ undertaken to make an overall assessment of the effectiveness of the 

Field test program. This asses