Test and Evaluation Trends and Costs For Aircraft and Guided Weapons

CHILD POLICY This PDF document was made available
CIVIL JUSTICE from www.rand.org as a public service of

EDUCATION
the RAND Corporation.
ENERGY AND ENVIRONMENT
HEALTH AND HEALTH CARE Jump down to document6

INTERNATIONAL AFFAIRS
NATIONAL SECURITY The RAND Corporation is a nonproﬁt

POPULATION AND AGING
research organization providing
PUBLIC SAFETY
SCIENCE AND TECHNOLOGY

objective analysis and effective
SUBSTANCE ABUSE solutions that address the challenges
TERRORISM AND
HOMELAND SECURITY
facing the public and private sectors
TRANSPORTATION AND around the world.
INFRASTRUCTURE
Support RAND
Purchase this document
Browse Books & Publications
Make a charitable contribution
For More Information

Visit RAND at www.rand.org
Explore RAND Project AIR FORCE
View document details
Limited Electronic Distribution Rights

This document and trademark(s) contained herein are protected by law
as indicated in a notice appearing later in this work. This electronic
representation of RAND intellectual property is provided for non-
commercial use only. Permission is required from RAND to reproduce, or
reuse in another form, any of our research documents.
This product is part of the RAND Corporation monograph series.
RAND monographs present major research ﬁndings that address the
challenges facing the public and private sectors. All RAND mono-
graphs undergo rigorous peer review to ensure high standards for
research quality and objectivity.
Test and Evaluation
Trends and Costs
for Aircraft and
Guided Weapons
Bernard Fox, Michael Boito, John C.Graser, Obaid Younossi
Prepared for the United States Air Force

Approved for public release, distribution unlimited
The research reported here was sponsored by the United States
Air Force under Contract F49642-01-C-0003. Further information
may be obtained from the Strategic Planning Division, Directorate of
Plans, HqUSAF.
Library of Congress Cataloging-in-Publication Data
Test and evaluation trends and costs for aircraft and guided weapons / Bernard Fox ...
[et al.].
p. cm.
“MG-109.”
Includes bibliographical references.
ISBN 0-8330-3540-1 (pbk. : alk. paper)
1. Airplanes, Military—United States—Testing. 2. Antiaircraft missiles—
United States—Testing. 3. Airplanes, Military—United States—Costs. 4. Antiaircraft
missiles—United States—Costs. 5. Air-to-surface missiles—Testing. 6. Air-to-surface
missiles—Costs. 7. United States. Air Force—Weapons systems—Testing.
8. United States. Air Force—Weapons systems—Costs. I. Fox, Bernard, 1951–
UG1243.T47 2004
358.4'18—dc22
2004005294
Cover photo courtesy of the U.S. Air Force at www.af.mil.

Photographer: Judson Brohmer
The RAND Corporation is a nonprofit research organization providing

objective analysis and effective solutions that address the challenges
facing the public and private sectors around the world. RAND’s
publications do not necessarily reflect the opinions of its research clients
and sponsors.
R® is a registered trademark.
© Copyright 2004 RAND Corporation
All rights reserved. No part of this book may be reproduced in any

form by any electronic or mechanical means (including photocopying,
recording, or information storage and retrieval) without permission in
writing from RAND.
Published 2004 by the RAND Corporation

1700 Main Street, P.O. Box 2138, Santa Monica, CA 90407-2138
1200 South Hayes Street, Arlington, VA 22202-5050
201 North Craig Street, Suite 202, Pittsburgh, PA 15213-1516
RAND URL: http://www.rand.org/
To order RAND documents or to obtain additional information, contact
Distribution Services: Telephone: (310) 451-7002;
Fax: (310) 451-6915; Email: [email protected]
Preface
This is one of a series of reports from a RAND Project AIR FORCE

project, “The Cost of Future Military Aircraft: Historical Cost Esti-
mating Relationships and Cost Reduction Initiatives.” The purpose
of the project is to improve the tools used to estimate the costs of
future weapon systems. It focuses on how recent technical, manage-
ment, and government policy changes affect cost.
This monograph examines the effects of changes in the test and
evaluation (T&E) process used to evaluate military aircraft and air-
launched guided weapons during their development programs.
Working from extensive discussions with government and industry
personnel, we characterize current trends in T&E and provide several
general estimating relationships that can be used early in program
development to estimate T&E costs. Appendixes A and B briefly
summarize relevant technical, schedule, and programmatic infor-
mation on recent test programs, while Appendix C provides official
definitions of the phases of T&E. A separate supplement provides
corresponding cost information but is available only to authorized
government personnel.
This project is being conducted within the RAND Project AIR
FORCE Resource Management Program. The research is sponsored
by the Principal Deputy, Office of the Assistant Secretary of the Air
Force (Acquisition), and by the Office of the Technical Director, Air
Force Cost Analysis Agency.
iii
iv Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
This monograph should be of interest to government cost ana-

lysts, the military aircraft and missile acquisition and T&E communi-
ties, and those concerned with current and future acquisition policies.
Other RAND Project AIR FORCE reports that address military
aircraft cost estimating issues include the following:
• An Overview of Acquisition Reform Cost Savings Estimates (Mark
Lorell and John C. Graser, MR-1329-AF) used relevant litera-
ture and interviews to determine whether estimates of the effi-
cacy of acquisition reform measures are robust enough to be of
predictive value.
• Military Airframe Acquisition Costs: The Effects of Lean
Manufacturing (Cynthia Cook and John C. Graser, MR-1325-
AF) examined the package of new tools and techniques known
as “lean production” to determine whether it would enable
aircraft manufacturers to produce new weapon systems at costs
below those predicted by historical cost-estimating models.
• Military Airframe Costs: The Effects of Advanced Materials and
Manufacturing Processes (Obaid Younossi, Michael Kennedy,
and John C. Graser, MR-1370-AF) examined cost estimating
methodologies and focus on military airframe materials and
manufacturing processes. This report provides cost estimators
with factors useful in adjusting and creating estimates based on
parametric cost estimating methods.
• Military Jet Engine Acquisition: Technology Basics and Cost-
Estimating Methodology (Obaid Younossi, Mark V. Arena,
Richard M. Moore, Mark Lorell, Joanna Mason, and John C.
Graser, MR-1596-AF) contains background information on
modern aircraft propulsion technologies and a variety of cost-
estimating methodologies.
RAND Project AIR FORCE

RAND Project AIR FORCE (PAF), a division of the RAND
Corporation, is the U.S. Air Force’s federally funded research and
development center for studies and analyses. PAF provides the Air
Preface v
Force with independent analyses of policy alternatives affecting the

development, employment, combat readiness, and support of current
and future aerospace forces. Research is performed in four programs:
Aerospace Force Development; Manpower, Personnel, and Training;
Resource Management; and Strategy and Doctrine.
Additional information about PAF is available on our Web site
at http://www.rand.org/paf.
Contents
Preface...............................................................................................................................................iii
Figures ..............................................................................................................................................xi
Tables............................................................................................................................................. xiii
Summary ........................................................................................................................................xv
Acknowledgments....................................................................................................................xxv
Abbreviations ........................................................................................................................... xxix
CHAPTER ONE
Background: The Cost of Testing ...................................................................................1
Changes in the Test Process ...................................................................................................5
T&E Constituencies...................................................................................................................6
Estimating the Cost of T&E..................................................................................................7
Study Objectives...........................................................................................................................9
Study Limitations ........................................................................................................................9
Organization of the Report.................................................................................................. 11
CHAPTER TWO
The T&E Process .................................................................................................................... 13
Types of Testing ........................................................................................................................ 16
Developmental Test and Evaluation ......................................................................... 16
Operational Test and Evaluation ................................................................................ 18
Multiservice Operational Test and Evaluation .................................................... 20
Qualification Operational Test and Evaluation .................................................. 20
Live-Fire Test and Evaluation ...................................................................................... 20
Test and Evaluation Process Steps.................................................................................... 21
vii
viii Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
Determine Test Objectives ............................................................................................ 22

Conduct Pretest Analysis ................................................................................................ 23
Conduct Test ........................................................................................................................ 23
Perform Post-Test Analysis ............................................................................................ 23
Report Results ...................................................................................................................... 24
Selected Test-Related Documentation........................................................................... 24
Analysis of Alternatives .................................................................................................... 25
Operational Requirements Document..................................................................... 25
Test and Evaluation Master Plan ................................................................................ 26
Test Resources ............................................................................................................................ 26
Modeling and Simulation............................................................................................... 27
Measurement Facilities..................................................................................................... 29
System Integration Laboratories.................................................................................. 29
Hardware-in-the-Loop Facilities ................................................................................. 29
Installed System Test Facilities .................................................................................... 29
Open-Air Ranges................................................................................................................. 30
CHAPTER THREE
Trends in Test and Evaluation ....................................................................................... 31
Acquisition Reform.................................................................................................................. 31
ORD Flexibility................................................................................................................... 32
Total System Performance Responsibility .............................................................. 34
Commercial Insertions (COTS/NDI) ..................................................................... 36
Modeling and Simulation..................................................................................................... 39
Testing of Software-Intensive Systems ........................................................................... 44
Combined and Multiservice Testing .............................................................................. 46
Contractor Versus Government Test Facilities.......................................................... 48
Live-Fire Testing ....................................................................................................................... 50
Warranties .................................................................................................................................... 51
Cost Trends ................................................................................................................................. 51
CHAPTER FOUR
Test and Evaluation Cost Data ...................................................................................... 59
Cost Element Definitions ..................................................................................................... 59
System Test and Evaluation .......................................................................................... 59
Developmental Test and Evaluation ......................................................................... 60
Contents ix
Operational Test and Evaluation ................................................................................ 60

Mock-Ups............................................................................................................................... 61
Test and Evaluation Support ........................................................................................ 61
Test Facilities ........................................................................................................................ 62
Cost Collection and Reporting .......................................................................................... 62
Contractor Costs ................................................................................................................. 63
Government Costs ............................................................................................................. 63
Cost Data Sources .................................................................................................................... 67
Cost Data Caveats .................................................................................................................... 68
How This Analysis Aggregated Cost Data ................................................................... 71
CHAPTER FIVE
Estimating Test and Evaluation Costs....................................................................... 75
Approaches to Estimating Test and Evaluation Costs ........................................... 75
A Priori Expectations of Aircraft ST&E Relationships.......................................... 80
Results of Aircraft ST&E Analysis ................................................................................... 82
Total Contractor ST&E.................................................................................................. 87
Contractor Ground Testing .......................................................................................... 89
Static and Fatigue Testing .............................................................................................. 90
Flight Testing........................................................................................................................ 91
Other Contractor Test Costs ........................................................................................ 95
A Priori Expectations for Guided-Weapon ST&E Relationships .................... 96
Results of Weapon ST&E Analysis ................................................................................. 99
Developing a T&E Cost Estimate................................................................................. 102
CHAPTER SIX
Conclusions and Recommendations........................................................................ 107
APPENDIX
A. Aircraft Program Descriptions ........................................................................... 111
B. Missile Program Descriptions ............................................................................ 145
C. System Test and Evaluation Work Breakdown Structure ................. 179
Bibliography ............................................................................................................................. 185

Figures
2.1. T&E Phases Within the Acquisition Process............................................ 14

2.2. System Maturity, Test Processes, and Resources .................................... 22
2.3. Number of Trials by Various Test Methodologies ................................ 27
3.1. Flight-Test Duration of Aircraft Development Programs Over
Time .............................................................................................................................. 52
3.2. Growth of Flight-Test Data Collected as Indicated by Number
of Sensors or Measurement Points ................................................................. 54
3.3. Comparison of Institutional and User Funding for Major Air
Force and Navy Aircraft and Weapon Test Centers ............................. 57
5.1. Total DT Flight Hours........................................................................................ 78
5.2. DT Flight Hours per Aircraft Month........................................................... 78
5.3. Guided DT Launches per Month in Order of Contract
Award Date ................................................................................................................ 79
5.4. Total Contractor ST&E Cost CER .............................................................. 87
5.5. Contractor Ground-Test CER......................................................................... 89
5.6. Static and Fatigue Test Costs............................................................................ 91
5.7. Contractor Flight-Test CER ............................................................................. 92
5.8. Percentage of Government Costs in Flight Test Have Been
Increasing .................................................................................................................... 94
5.9. CER for Other Contractor Test Costs......................................................... 95
5.10. Missile Total ST&E CER (Excludes Guided Munitions) .............. 101
xi
Tables
1.1. T&E Perspectives.......................................................................................................6

2.1. Primary Air Force and Navy Open Air Ranges for Aircraft and
Missile Testing ......................................................................................................... 30
3.1. How Advanced Capabilities Affect Testing............................................... 54
3.2. Comparison of F/A-18 Fatigue Testing ...................................................... 56
4.1. Representative Contractor DT&E Activities for Aircraft and
Guided Weapons .................................................................................................... 61
4.2. Aircraft Cost Data Sources................................................................................. 67
4.3. Guided-Weapon Cost Data Sources ............................................................. 68
5.1. Average Test Costs for Representative Multiengine Fighters ........... 77
5.2. Summary of Aircraft Test Program Characteristics ............................... 81
5.3. Definitions of Aircraft Variables ..................................................................... 83
5.4. Aircraft Variables Correlation Matrix (11 Programs) ........................... 86
5.5. Summary of T&E Estimating Resources .................................................... 87
5.6. Total ST&E CER—Preferred.......................................................................... 88
5.7. Total ST&E CER—Alternative 1 .................................................................. 88
5.8. Total ST&E CER—Alternative 2 .................................................................. 88
5.9. Ground-Test CER—Preferred......................................................................... 90
5.10. Ground-Test CER—Alternative..................................................................... 90
5.11. Static and Fatigue Test CER—Preferred .................................................... 91
5.12. Flight-Test CER—Preferred ............................................................................. 93
5.13. Flight-Test CER—Alternative 1 ..................................................................... 93
5.14. Flight-Test CER—Alternative 2 ..................................................................... 93
5.15. Other Contractor Test CER—Preferred .................................................... 96
xiii
xiv Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
5.16. Summary Characteristics of Missiles and Guided

Munitions ................................................................................................................... 97
5.17. Variable Definitions for Missiles .................................................................. 100
5.18. Missile Variables Correlation Matrix (10 Programs) ......................... 100
5.19. Missile Total ST&E CER—Preferred ...................................................... 101
5.20. Missile Total ST&E CER—Alternative................................................... 102
5.21. Indicators of Test Program Scope ............................................................... 103
A.1. B-1B Flight Testing ............................................................................................ 113
A.2. The B-1B Test Aircraft ..................................................................................... 113
A.3. B-2 Flight-Test Program .................................................................................. 116
A.4. The B-2 Test Aircraft......................................................................................... 116
A.5. C-17 Flight-Test Program............................................................................... 120
A.6. The C-17 Aircraft Tested................................................................................. 120
A.7. F/A-18E/F Flight-Test Program .................................................................. 124
A.8. The F/A-18E/F Test Aircraft ......................................................................... 124
A.9. Special Requirements for F/A-22 Testing ............................................... 129
A.10. F/A-22 Flight Test; June 2001 Replan ..................................................... 131
A.11. The F/A-22 Test Aircraft ................................................................................. 131
A.12. F/A-22 Planned Flight-Test Program........................................................ 132
A.13. T-45 Flight Testing............................................................................................. 135
A.14. T-45 Testing Summary..................................................................................... 136
A.15. T-45C Flight Testing ......................................................................................... 137
A.16. V-22 Flight Testing............................................................................................. 142
A.17. V-22 Testing Summary..................................................................................... 143
B.1. AIM-9X Testing ................................................................................................... 148
B.2. AMRAAM Testing ............................................................................................. 155
B.3. JASSM Testing...................................................................................................... 162
B.4. JDAM Testing....................................................................................................... 166
B.5. JSOW Testing ....................................................................................................... 170
B.6. SFW Testing .......................................................................................................... 173
B.7. SLAM-ER Testing............................................................................................... 175
B.8. WCMD Testing ................................................................................................... 177
Summary
T&E is a key step in the development of any military weapon system.

It is the primary means of ensuring that the system will actually per-
form its intended functions in its intended environment.
T&E of a modern weapon system is an involved and often
lengthy process that reflects both the complexity of the system under
test and the variety of specialized resources and activities its testing
requires. T&E consumes a significant portion of the development
time and resources for military aircraft and air-launched weapons,1
which is why the general reexamination of acquisition processes that
has taken place over the past decade has included T&E. Looking for
efficiencies and cost savings, advocates of acquisition streamlining
have questioned the scope, duration, cost, and organizational respon-
sibilities of the traditional T&E process. These questions are even
more urgent because most T&E expenditures occur in the later stages
of development, when cost overruns and schedule slips from other
activities may have become more apparent. As a result, there is often
considerable pressure to expedite and/or reduce T&E activities to
recoup some of the other overruns.
The T&E process has evolved with the complexity and cost of
the systems being developed and with the priorities and practices of
defense acquisition management. This evolution and its effects on the
development cost of the systems under test are the subject of this
monograph.
_____________
1On average, contractor and government T&E account for approximately 21 percent of
development costs for fixed-wing aircraft and 15 percent for guided weapons.
xv
xvi Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
The tasking for this study arose from two concerns. Some pro-
gram managers have proposed test programs of greatly reduced scope
and duration, citing such initiatives as increased use of modeling and
simulation to reduce the amount of expensive “open air” testing.
Other rationales for reduced test schedules and budgets include using
lower-risk designs, combining government and contractor testing,
using nondevelopmental item (NDI) and commercial-off-the-shelf
(COTS) approaches, and applying total system performance respon-
sibility (TSPR) contracting. Acquisition decisionmakers needed to
know whether these approaches can achieve the projected savings.
The second concern was that members of the cost analysis
community, particularly those outside of the program offices and test
organizations, were not confident that the data and relationships they
were using to estimate the costs of testing for a program or to cross
check such estimates reflected the current T&E environment. Since
some of their tools were based on development programs that were
15 to 30 years old, validation against current and evolving T&E
approaches became a priority.
Although the original intention was for this study to focus on
fixed-wing aircraft, the Air Force Cost Analysis Agency (AFCAA)
asked RAND Corporation to include a cross section of tactical mis-
siles and guided munitions. Because many of the programs of interest
were joint Air Force–Navy development efforts and because the Navy
cost community had similar requirements, the Assistant Secretary of
the Navy for Research, Development, and Acquisition (ASN RDA)
agreed and directed the appropriate Navy program executive officers
and test activities to support the project.
The project scope involved the following four tasks:
• analyzing the nature of current T&E costs for aircraft, tactical
missile, and guided munition systems and the trends likely to
affect these costs in the immediate future
• identifying key cost drivers
• collecting, normalizing, and documenting representative data
• developing a set of practical, documented methodologies for
making high-level T&E estimates.
Summary xvii
To interpret the results of this study correctly, certain limita-

tions and constraints should be kept in mind. First, the study focused
on system-level testing associated with development programs funded
through research, development, test, and evaluation and categorized
as “system T&E.” This therefore excluded postproduction follow-on
testing, production acceptance testing, and component-level testing.
Second, the study focused only on what government program
offices typically pay for, the items test organizations often refer to as
reimbursable costs. These could be considered the price the customer
pays for test services. These T&E costs are part of each weapon sys-
tem’s development budget, whether it is the contractor or the pro-
gram office that directly incurs them.2
Third, we limited our analysis to recent Air Force and Navy
fixed-wing aircraft, tactical missile, and guided munition programs.
Because the purpose of the study was to examine current test prac-
tices, we focused generally on programs that had completed develop-
ment within the past ten years or, in a few cases, slightly earlier.3
Older data were used for trend analysis and, where appropriate, to
augment more-recent data in developing relationships.
Because fewer new development programs are projected for the
future, we attempted to include several programs representing major
modifications to existing systems for which enough data were avail-
able for our analysis. Relevance to both modification and new devel-
opment programs was also a consideration in selecting parameters for
cost relationships.
Since our purpose was to examine the cost of testing as it was
being conducted at the time of our research, we did not assess the
efficiency and effectiveness of test procedures.
_____________
2 The government does pay other T&E expenses, such as overhead and construction at test
facilities, through specific appropriations. These are not allocated to any weapon system and,
therefore, are not included in this study.
3 One of the purposes of the study was to provide more-current cost, technical and
programmatic data to the cost community. In a few cases, the data we collected were slightly
older than our nominal 10 years but were not always generally available within the cost
organizations and thus would be a useful resource.
xviii Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
Cost-Estimating Data, Methodologies, and Trends

To develop cost-estimating methodologies, we collected actual T&E
costs, schedules, and programmatic test data from a number of
sources, including the contractor cost data reports, system program
offices, government cost analysis agencies, government test organiza-
tions, and selected contractors (see Acknowledgments). Chapter Four
discusses these data, which we treat more fully in a limited-
distribution supplement. The appendixes include detailed program-
matic data on 16 aircraft and guided-weapon programs for reference.
Chapter Five presents the T&E cost estimating relationships
(CERs) we developed from these data. The CERs and the data in the
proprietary supplement should allow an estimator to compare esti-
mates for a proposed program with actuals from other programs. Of
course, the estimator will have to use expert judgment to take into
account any specific, unique aspects of the proposed program. Chap-
ter Five includes CERs for
• overall contractor test costs for aircraft
• contractor ground, flight, and “other” test costs for aircraft
• total contractor and government test costs for guided weapons.
As with most parametric estimating tools, these would be most
useful for a Milestone B or earlier cost or test estimate, when fewer
details of a proposed program are known. As the system progresses
through development and more information becomes available,
more-detailed estimating techniques can be used, with these CERs
providing a cross-check at an aggregate level.
It was much more difficult to collect and document data on the
costs government organizations had incurred than on corresponding
contractor costs. We initially did not consider this to be a serious
limitation, assuming that, because of acquisition reform, government
costs would decrease as contractors took on a greater share of the
effort. However, in cases where we were able to obtain government
costs for programs, this generally did not prove true. Government
T&E costs were substantial and, for guided weapons, generally
greater than those of the system contractor. In many cases, contrac-
Summary xix
tors still rely on government test facilities and functional expertise,

particularly for high-cost, low-utilization test capabilities. Govern-
ment personnel normally participate actively in the integrated test
teams. Even when the contractor can select any test facility it consid-
ers appropriate, that might end up being a government facility—with
the government facility then becoming test subcontractor to the sys-
tem prime contractor. Of course, most open-air testing continues to
be conducted on DoD ranges.
Consistent accumulation and reporting of government cost data,
to standards similar to those for contractor data, would greatly
improve the accuracy of cost estimates for testing. This would ensure
that the total program financial picture was available for management
in the present and for analysis in the future. This would improve the
ability of government test facilities to evaluate the cost and schedule
implications of their processes, assess the contributions of all their
activities, and focus investment and management attention on the
activities most critical to each facility’s customer base.
Overall T&E Cost Trends

The overall cost of T&E to the program shows no clear trend upward
or downward over the last 20 to 30 years. Although government and
industry test personnel have indicated that the increasing use of mod-
eling and simulation, improvements in instrumentation and test
processes, reduction of redundant testing, and various acquisition
streamlining initiatives have reduced the cost of individual tests, other
changes appear to have offset any potential net savings.
Thus, the proportion of development costs dedicated to T&E
has remained relatively constant for aircraft and guided weapon sys-
tems. Although various explanations for this are possible, the domi-
nant factors are probably the increasing complexity of the systems
tested and the increasing content of test programs. (See the Cost
Trends section in Chapter Three.)
xx Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
T&E Issues and Findings

Another principal objective of this study was to identify changes in
the practice of T&E and, to the extent possible, their likely effects on
the cost of T&E for future aircraft, missiles, and guided munitions.
Overall, we found no cost or schedule data that would allow us
to quantify how these practices individually affect current systems,
either as upward or downward influences on test costs or schedules.
The following paragraphs outline the issues we addressed.
Acquisition Reform
Acquisition reform initiatives are a diverse array of ideas, processes,
and practices designed to streamline the DoD acquisition process,
reducing either cost or schedule, or improving technology. A previous
RAND report (Lorell, 2001) addressed the general effects of
acquisition reform on cost estimating.
One of the acquisition reform initiatives that report discusses is
TSPR, which transfers certain T&E responsibilities from DoD to the
contractors. Although the data to support cost savings tend to be
anecdotal, it is apparent that it will shift costs from government
activities to contractor activities and must be recognized as such in
future cost estimates. Our interviews suggest that TSPR must be well
planned to avoid two test-related problems: Test data may not be
available to DoD for other, postdevelopment uses, and cross-platform
integration might not be adequately coordinated, especially in guided
weapon development. DoD must have the foresight to ensure that it
can use the system development and demonstration test data to
design modifications or to qualify additional platform-and-weapon
configurations. In addition, to maintain compatibility, DoD will
have to ensure careful coordination of cross-platform integration
issues, particularly with guided-weapon development and modifica-
tion, with other systems.
It is too early to assess the outcome of recent innovative test
management approaches that give the contractor broad latitude in
developing and executing the developmental test program. Another
innovative approach, relying on non-DoD tests and certifications of
Summary xxi
nondevelopmental aircraft for DoD applications, was not generally as

successful as its promoters had hoped. We found that Federal Avia-
tion Administration certification alone is not sufficient to demon-
strate that a particular aircraft meets most military performance speci-
fications. “Best commercial practices” are not an effectively codified
set of procedures, like common law or accounting principles. Because
they tend to be situational and inconsistent from contractor to con-
tractor, they may be inadequate for responsible acceptance of military
systems. (See the Acquisition Reform section in Chapter Three.)
Modeling and Simulation

Virtually all test programs now incorporate modeling and simulation.
In many programs, some aspects of the analytical tools have not been
mature enough to give enough confidence for waiving live testing.
However, in all cases, modeling and simulation at least reduced the
risk, and often the duration, of live tests and thus appeared to be a
good investment. In addition to directly benefiting T&E, robust
modeling and simulation also benefits
• evaluating design excursions during development
• tactics development
• operator training
• evaluating future system enhancements. (See the Modeling and
Simulation section in Chapter Three.)
Testing of Software-Intensive Systems

An area of almost universal concern was effective testing of software
intensive systems, which are growing in complexity and functionality.
Continuing advances in technology have translated into system capa-
bilities unimagined a generation ago. The growth in capability trans-
lates into increased test complexity. This area should receive specific
attention in any future T&E estimates. (See the Software Intensive
Systems section in Chapter Three.)
xxii Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
Combined and Multiservice Testing

There was general agreement that integrated contractor-government
test teams were a positive force in optimizing testing. Similarly, com-
bined development and operational test teams have been valuable
because they avoid redundant testing and highlight operational effec-
tiveness and suitability issues for early resolution. Some program per-
sonnel expressed a desire for even more intensive “early involvement”
by the operational test community. The primary constraint appears to
be limited staffing of the service operational test organizations. (See
the Combined and Multiservice Testing section in Chapter Three.)
Contractor Versus Government Test Facilities

While there was general agreement that the major government test
facilities are essential for executing the required test programs and
that they generally provide excellent support, some contractor per-
sonnel expressed varying levels of frustration in their dealings with the
government test organizations. In programs with aggressive afforda-
bility goals, there was a concern that some government test range per-
sonnel were not as focused on controlling the costs and schedule of
the test program as other members of the test team were. Some felt
that there were practices at the ranges that were overly conservative
and caused unnecessary costs and delays. In other cases, delays
resulted from chronic understaffing or procedures with little
provision for flexibility. These issues are of increasing importance
when contractors are given incentives to perform within what are, in
effect, fixed test budgets and schedules. A related contractor concern
was that the government ranges tended to be “overfacilitized” but
“undermodernized.” (See the Contractor Versus Government Test
Facilities section in Chapter Three.)
Live-Fire Testing
Although live-fire testing can be a contentious issue during early
planning for system development, our interviews did not highlight
major concerns at the program level, as long as the requirements were
known in advance and planned for accordingly. Because data were
Summary xxiii
limited, we could draw no general conclusions about the real cost of

live-fire testing. (See the Live-Fire Testing section in Chapter Three.)
Although there is some disagreement over the appropriate level
of testing in specific circumstances—live-fire testing, testing for statis-
tically rare events, etc.—we found little controversy in general over
the scope of testing. Other studies have concluded that most DoD
test programs have already eliminated the majority of unnecessary or
redundant testing. Several sources, however, expressed the opinion
that thoughtful reevaluation of test procedures could improve the
pace and efficiency of the typical test program. (See the Live-Fire
Testing and the Contractor Versus Government Facilities sections in
Chapter Three.)
Warranties
None of our interviews indicated that warranties significantly
changed the T&E process or costs.
Acknowledgments
In performing this study, the authors had extensive discussions with

subject-matter experts in both government and industry. Indeed,
without their willingness to share their time, insights, and data gener-
ously, this study would have been more limited in both scope and
utility. To all who so willingly supported this effort, we offer our sin-
cere thanks. Unfortunately, they are far too numerous to mention
individually, but we would like to specifically acknowledge the prin-
cipal organizations visited and our key points of contact in each:
• Air Force Cost Analysis Agency: Joseph Kammerer, Director;
Jay Jordan, Technical Director
• Director of Test & Evaluation, Headquarters, USAF (AF/TE):
John T. Manclark
• Office of the Secretary of Defense (OSD/PA&E): Gary Bliss
• Director, Operational T&E: Christine Crabill
• 46th Test Wing: Carlos Rodgers
• Aeronautical Systems Center: Michael Seibel
• Air Armament Center: Jung Leong
• Air Force Flight Test Center: James Dodson; Vicky Yoshida
• Air Force Operational Test Center: Joseph Guy
• Chief of Naval Operations (N091): Skip Buchanan
• Commander, Operational T&E Force: Commander William
Padgett
• Naval Air Systems Command: David Heller (AIR-5.0); David
Volpe (AIR-4.2)
xxv
xxvi Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
• Naval Air Warfare Center—Aircraft Division: J. R. Smullen;

Robert Mann
• Naval Air Warfare Center—Weapons Division: Robert Ostrom;
Robert Copeland
• Naval Center for Cost Analysis: Bill Stranges
• B-1B: Rick Wysong
• B-2: Larry Perlee
• C-17: Eric Wilson
• F-16: Douglas Egged
• F-18: Steve Kapinos
• F-22: Melanie Marshall; Mark Whetstone
• JPATS: Jay Free
• Joint Strike Fighter: Paul Tetrault
• T-45: Lieutenant Commander Jennifer Rigdon
• V-22: Sandie Raley
• AIM-9X: Toby Jones
• AMRAAM: Robert Guidry; Jack Trossbach
• JASSM: Commander Patrick Roesch
• JDAM: Mike Evans
• JSOW: Michael Chartier
• Sensor Fuzed Weapon: Beth Crimmins
• SLAM-ER: Carl Smith
• Tomahawk: Lieutenant Commander Tim Morey
• WCMD: Duane Strickland
• Boeing Military Aircraft and Missile Systems: Kurt Syberg
• Lockheed Martin Aeronautics: Paul Metz
• Raytheon Missile Systems: Glen Pierson.
We would also like to thank several individuals in particular
whose support was both extensive and central to the completion of
this study: Jay Jordan for his guidance and helpful suggestions; Dave
Heller, whose encouragement and persistence opened many Navy
doors; Bill Stranges, who provided both thoughtful feedback and
missing data; and Steve Cricchi, Joe Guy, Bob Kellock, CDR Bill
Padgett, Larry Perlee, Kurt Syberg, Paul Tetrault, Jack Trossbach, Joe
Acknowledgments xxvii
Wascavage, Eric Wilson, and Vicky Yoshida, whose insights and

efforts in locating key data were indispensable.
Our RAND colleagues Michael Kennedy and Dave Stem
improved the report significantly by their thorough review and con-
structive comments. Tom Sullivan assisted with the data analysis.
Michele Anandappa provided invaluable administrative support.
Abbreviations
AAC Air Armament Center

AAV AMRAAM Air Vehicle
AAVI AMRAAM Air Vehicle Instrumented
ACAT acquisition category
ACE AMRAAM captive equipment
ACTD advanced concept technology demonstra-
tion
AEDC Arnold Engineering Development Center
AFB air force base
AFCAA Air Force Cost Analysis Agency
AFMSS Air Force Mission Support System
AFOTEC Air Force Operational Test and Evaluation
Center
AIAA American Institute of Aeronautics and
Astronautics
AMRAAM Advanced Medium-Range Air-to-Air Mis-
sile
AoA analysis of alternatives
APREP AMRAAM Producibility Enhancement
Program
ASN (RDA) Assistant Secretary of the Navy for
Research, Development, and Acquisition
BAe British Aerospace
xxix
xxx Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
CAIV cost as an independent variable

CARD cost analysis requirements description
CBU cluster bomb unit
CCDR contractor cost data report
CCRP captive-carry reliability program
CDR critical design review
CER cost estimating relationship
CLV captive-load vehicle
CMUP Conventional Mission Upgrade Program
COI critical operational issue
COMOPTEVFOR Commander, Operational Test and
Evaluation Force
COTS commercial-off-the-shelf
CPR cost performance report
CTD concept and technology development
CTF combined test force
CTU captive-test unit
DEM/VAL demonstration and validation
DoD Department of Defense
DOT&E Director of Operational Test and Evalua-
tion
DT developmental testing
DT&E development test and evaluation
EAC estimate at completion
EDM engineering development missile
EMC electromagnetic compatibility
EMD engineering and manufacturing develop-
ment
EMI electromagnetic interference
EW electronic warfare
Abbreviations xxxi
FAA Federal Aviation Administration

FDE force development evaluation
FOT&E follow-on operational test and evaluation
FMS foreign military sales
FSD full-scale development
FTTC Flight Test Technical Committee
FY fiscal year
GAM GPS-aided munition
GPS Global Positioning System
GTV guided test vehicle
HITL hardware in the loop
ILS instrument landing system
IMV instrumented measurement vehicle
INS inertial navigation system
IOC initial operational capability
IOT&E initial operational test and evaluation
IPT integrated product team
IR infrared
IRT instrumented round with telemetry
ISTF installed system test facility
IT integrated testing
JASSM Joint Air-to-Surface Standoff Missile
JDAM Joint Direct Attack Munition
JET joint estimating team
JHMCS Joint Helmet Mounted Cueing System
JPATS Joint Primary Aircraft Training System
JSF Joint Strike Fighter
JSOW Joint Standoff Weapon
LFT&E live-fire test and evaluation
xxxii Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
LO low observability
LRIP low-rate initial production
M&S modeling and simulation
Mk. mark
MNS mission needs statement
MOE measure of effectiveness
MOT&E multiservice operational test and evaluation
MRTFB Major Range and Test Facility Base
NAVAIR Naval Air Systems Command
NAWC-AD Naval Air Warfare Center–Aircraft Division
NAWC-PM Naval Air Warfare Center–Pt. Mugu
NDI nondevelopmental item
OAR open-air range
OPEVAL operational evaluation
OPTEVFOR Operational Test and Evaluation Force
ORD operational requirements document
OSD Office of the Secretary of Defense
OT operational testing
OTA Operational Test Agency
OT&E operational test and evaluation
P3I preplanned product improvement
PD/RR program definition/risk reduction
PEP Producibility Enhancement Program
PID program introduction document
PRTV production representative test vehicle
QOT&E qualification operational test and evaluation
QT&E qualification test and evaluation
RCS radar cross section
RDT&E research, development, test, and evaluation
Abbreviations xxxiii
SAF/AQ Assistant Secretary of the Air Force (Acqui-

sition)
SCTV separation control test vehicle
SDD system development and demonstration
SFW sensor fuzed weapon
SLAM-ER Standoff Land-Attack Missile–Expanded
Response
SOC statement of capability
SPO system program office
STARS Standard Accounting and Reporting System
ST&E system test and evaluation
STU seeker test unit
STV separation test vehicle
T&E test and evaluation
T1 theoretical first unit
TAMPS Tactical Automated Mission Planning Sys-
tem
TASSM Triservice Air-to-Surface Attack Missile
TECHEVAL technical evaluation
TEMP test and evaluation master plan
TSPR total system performance responsibility
UJFT Undergraduate Jet Flight Training System
USC United States Code
VCD verification of correction of deficiencies
VOR very-high-frequency omnidirectional range
(station)
WBS work breakdown structure
WCMD Wind-Corrected Munitions Dispenser
WIPT working-level integrated product team
CHAPTER ONE
Background: The Cost of Testing
Test and evaluation (T&E) is a key step in the development of any

military system. It is the primary means of ensuring that the system
will actually perform its intended functions in its intended environ-
ment.
T&E of a modern weapon system is an involved and often
lengthy process reflecting both the complexity of the system under
test and the variety of specialized resources and activities required. As
Figures 1.1 through 1.4 show, for military aircraft and air-launched
weapons, T&E represents a significant portion of development
resources and time.1 Because of this, T&E has been included in the
general reexamination and attempts to improve acquisition processes
conducted over the past decade. Looking for efficiencies and cost
savings, advocates of acquisition streamlining have questioned the
scope, duration, cost, and organization of the traditional T&E
process. These questions are even more urgent because most T&E
expenditures occur in the later stages of development, when cost
overruns and schedule slips from other activities may have already
occurred. As a result, although T&E may or may not be the pacing
activity in a program’s execution, there is often considerable pressure
to reduce T&E activities to compensate for losses due to overruns and
slips.
_____________
1 To keep Figures 1.1 and 1.2 nonproprietary but to give the reader a sense of the magnitude of the test
costs, we divided total contractor and government test costs by the total full-scale development (FSD)
and engineering and manufacturing development (EMD) costs for each system.
1
2 Test and Evaluation Trends and Costs for Aircraft and Guided Weapons
Figure 1.1
Aircraft System T&E Share of Total FSD and EMD Costs
30
Government
Contractor
25
Percentage of total cost
20
15
10
0
F/A-18E/F F-22* B-2 B-1 CMUP C-17 T-45 V-22* Avg.
BLK D
NOTE: * Program in progress
RAND MG109-1.1
Figure 1.2
Guided-Weapon System T&E Share of Total FSD and EMD Cost
25
Government
Contractor
20
Percentage of total cost
15
10
0
AMRAAM AMRAAM AIM-9X JASSM* SFW JDAM JSOW Avg.
FSD P3I Baseline

RAND MG109-1.2
Background: The Cost of Testing 3
Figure 1.3
DT and OT Share of Total Aircraft FSD and EMD Duration
200
DT start to OT end
MSII to DT start
150
Test duration (months)
100
50
0
F-18A/B F-18E/F F-22* B-2 B-1 C-17 V-22 T-45 Avg.
CMUP EMD*
RAND MG109-1.3
Figure 1.4
DT and OT Share of Total Guided-Weapon FSD and EMD Duration
90
DT start to OT end
80
MSII to DT start
70
Test duration (months)
60
50
40
30
20
10
0
AMRAAM AIM-9X SFW JDAM JSOW Avg.
Baseline
RAND MG109-1.4
The T&E process has evolved with the complexity and cost of the
systems being developed and with the priorities and practices of
defense acquisition management. This evolution of the T&E process
and its effect on the development cost of the systems under test are
the subject of this monograph.
The tasking for this study arose from two concerns. The first
was the desire of senior acquisition managers to get a satisfactory
answer to a seemingly straightforward question: “Are we budgeting an
appropriate amount for T&E in this program?” With the pervasive
emphasis on acquisition streamlining, commercial practices, cost-as-
an-independent-variable (CAIV) approaches, and a variety of lean
initiatives, there was a concern that the current T&E process might
be a holdover from when contractors built hardware to detailed gov-
ernment specifications and standards and when redundant, stove-
piped processes increased both cost and schedule.
On the other hand, program managers, under intense afforda-
bility pressures, were reexamining every aspect of their programs and
looking for potential cost savings. As a result, some program manag-
ers were proposing to reduce the scope and duration of test programs
greatly, citing such initiatives as increased use of modeling and simu-
lation (M&S) to reduce the amount of expensive “open air” testing.
Other rationales for reduced test schedules and budgets include using
lower-risk designs, combining government and contractor testing,
using nondevelopmental item (NDI) and commercial-off-the-shelf
(COTS) approaches, and applying total system performance respon-
sibility (TSPR) contracting.
The second concern was that members of the cost analysis
community, particularly those outside of the program offices and test
organizations, were not confident that the data and relationships they
were using to provide early estimates of the costs of testing for a pro-
gram or to cross check such estimates reflected current trends in the
current T&E environment. Since some of their tools were based on
development programs that were 15 to 30 years old, validation
against current and evolving T&E approaches was a priority.
Changes in the Test Process

Independent of specific budget and schedule pressures, the approach
to T&E should be consistent with the streamlined approaches that
have been introduced into other areas of defense acquisition man-
agement.
A major thrust of acquisition reform has been to integrate and
streamline what were perceived as stovepiped processes, to reduce the
time and cost they involved. In particular, such initiatives as inte-
grated product teams (IPTs) were established to make planning and
execution of tests more efficient, to improve communications among
the various disciplines and organizations involved, and to improve
their ability to respond to unplanned events. In T&E, this begins
with combining contractor test personnel and their government
counterparts into an integrated developmental testing (DT) team.
The intent is to ensure that the contractor, who generally has lead
responsibility for executing the DT program, meets the government
DT data requirements with sufficient visibility during tests to allow
the program to progress. This parallels the IPT organization in other
areas of defense acquisition.
A second integration initiative was the formation of combined
test forces (CTFs). These allow the operational testing (OT) commu-
nity to participate substantially in DT activities that formerly focused
on proving engineering designs rather than operational effectiveness.
This has two desirable results:
• Involving the OT community helps identify potential opera-
tional effectiveness and suitability issues early, when corrective
measures are generally less disruptive and less expensive.
• Maximizing the ability of the OT community to use data devel-
oped during DT has the potential to reduce or eliminate redun-
dant dedicated OT.
Another trend in defense acquisition that has affected T&E has
been the use of a reduced number of top-level performance specifica-
tions rather than detailed item specifications. This requires the test
community to design test programs that both demonstrate the
achievement of critical performance parameters and minimize the

time and resources needed to do so.
Trends in technology have had the most obvious effects on the
way T&E is conducted today. The systems under test are considera-
bly more complex than those even one generation older. This has
required a corresponding improvement in the capabilities of testing
tools and methods. The increasing use of M&S is an obvious example
of how technology has improved testing capability and productivity.
What may not be so obvious is that these tools are often necessary
because live testing of all functions of current systems is impractical,
if not impossible.
T&E Constituencies
Different constituencies bring diverse perspectives to T&E (see Table
1.1). Depending on the program phase and how the data are to be
used, T&E can be either a learning process or the final exam. Opti-
mizing T&E for each constituency would tend to drive T&E priori-
ties and objectives in different directions. For example, system
designers tend to view T&E as an integral part of the development
process. Depending on the type of test, designers may view it as an
experimental confirmation of the engineering design approaches or a
realistic exercise of a complex system to discover incompatibilities or
integration problems. In most cases, designers and engineers
approach T&E as a learning process; the more rigorous the test, the
more you learn, but the more likely you are to fail. This also tends to
be the approach in commercial product testing because “in-service”
product failures can adversely affect future sales.
Table 1.1
T&E Perspectives
Role T&E Objective

Designer Insight into expected versus actual performance
Manager Evidence of progress, design maturity, risk reduction
User Assurance that the system can perform its intended mission
Managers, who are concerned with schedules and budgets, tend

to view T&E as a series of milestones that signify graduation from
one phase of development to another. Rigorous T&E consumes
resources and time and must therefore be carefully structured,
managed, and optimized so standards are met without overtesting.
From the management perspective, failure can lead to costly redesign
and retesting. One comment that the Director of Operational Test
and Evaluation (DOT&E) has heard is that a “dollar spent on T&E
is a dollar spent looking for trouble” (Jonson, 2002). Yet the Defense
Science Board Task Force on T&E found that“[i]n most DoD
[Department of Defense] test programs, the content is already at or
near minimum” (Defense Science Board, 1999).
The ultimate users of the system tend to view T&E (especially
OT) as their assurance that the system will actually perform its mis-
sion in the “real world.” Theirs is the concern that a system that per-
forms its functions in a limited, possibly idealized environment may
be purchased and fielded yet fail when subjected to operation and
maintenance under field conditions. While designers and managers
tend to focus on compliance with technical specifications, operational
testers are primarily concerned with end-to-end testing against the
ultimate operational requirements.
Estimating the Cost of T&E

The cost of a T&E program can be estimated in several ways. The
specific approach taken is normally a function of the time and data
available to do the cost estimate, as well as the organization directing
the estimate.
The initial T&E estimate is generally developed as part of the
planning budget for the development program. As in many estimat-
ing situations, initial estimates are often based on comparisons with
the actual cost of similar predecessor programs. This approach
depends on the timely availability of cost and characteristic data of
the analogous programs. If the programs are similar overall, the cost
data may be adjusted in those areas where the similarities are not as
strong to better represent the characteristics, and presumably the cost,

of the program being estimated.
A second approach to estimating the cost of T&E for a new sys-
tem is parametric analysis. Analysts collect cost data from multiple
programs and compare them through statistical tests to uncover cost-
driving program characteristics. This approach requires not only that
historical cost, technical, and programmatic data be available for mul-
tiple programs but also that the data set either be homogeneous
enough or have enough data points to permit statistical stratification
by class. The simplest parametric relationships represent T&E cost as
a constant multiple of a single cost-driving parameter (for example,
development cost, number of flight-test hours). Given sufficient data,
analysts can develop more-robust cost-estimating relationships
(CERs) statistically using multiple parameters.
The third approach is to sum lower-level costs estimated by
various methods. This method is typically used after test plans have
become available and once the types and quantities of test resources
have been identified. Primary test organizations typically do this type
of estimating, working from the test requirements system program
office (SPO) has provided. Normally the estimator uses cost rates per
test resource unit (for example, number of flight-test hours, type of
data collection, target or threat presentations) to estimate the cost for
the testing required. Test organizations typically set their rates annu-
ally. This method requires accurate descriptions of the program
requirements and estimators with the experience to make allowances
for the inevitable scrubs, delays, and retests. This detailed, build-up
approach has the advantages of accounting for the unique require-
ments of a particular program and of being useful for setting and
tracking budgets. Its disadvantage is that it is based on assessments of
future test requirements, which may not account for contingencies or
may otherwise underscope effort or content.
Study Objectives
The Office of the Assistant Secretary of the Air Force (Acquisition)
(SAF/AQ) tasked RAND Project AIR FORCE to analyze the major
categories of acquisition support costs (commonly referred to as
“below-the-line costs”) to improve the Air Force cost-analysis com-
munity’s estimating approaches and tools. Discussions with the pro-
ject technical monitor at the Air Force Cost Analysis Agency
(AFCAA) indicated that the most pressing need was to examine sys-
tem T&E costs. Although the original intention was to focus on
fixed-wing aircraft, AFCAA asked RAND to include a cross section of
tactical missiles and guided munitions as well. Since many of these
programs were joint Air Force–Navy development efforts and since
the Navy cost community had similar requirements, SAF/AQ
requested Navy support and participation through the Office of the
Assistant Secretary of the Navy for Research, Development, and
Acquisition (ASN [RDA]). ASN (RDA) agreed and directed the
appropriate Navy program executive officers and test activities to
support the project.
The project involved the following four tasks:
• analyzing the nature of current aircraft, tactical missile and
guided munition system T&E costs and trends likely to affect
them in the immediate future
• identifying key cost drivers
• collecting, normalizing, and documenting representative data
• developing a set of practical and documented cost-estimating
methodologies.
Study Limitations
To interpret the results of this study correctly, certain limitations and
constraints should be kept in mind. First, the study includes only
those costs typically paid for by the program offices (which the test
organizations often refer to as reimbursable costs). These costs could be
considered the “price to the customer” for test services. Specifically,
we did not collect or analyze cost data on the internal costs or opera-
tions of the DoD test infrastructure.2 The program-funded test costs
are the ones that are included in weapon system estimates prepared
for service and Office of the Secretary of Defense (OSD) reviews and
that are included in program budgets. Direct appropriations cover the
costs of test organizations’ facilities and infrastructure rather than the
“variable” costs of testing. A variety of studies and panels have been
examining the broader issues of test infrastructure.
We limited the programs we analyzed to recent Air Force–Navy
fixed-wing aircraft, tactical missiles, and guided munitions. We
focused on programs that had completed development within the
past ten years or, in a few cases, slightly earlier, if test costs and pro-
grammatic data for those programs were not readily available to DoD
cost analysts.3 Older data from existing reports was used for trend
analysis and, where appropriate, to augment more-recent data in
developing relationships.
In an effort to consistently define testing to be included in the
study, we also limited ourselves to the testing categorized as system
T&E and to that associated with RDT&E-funded development
programs. Thus, the study results do not capture subsystem testing
(such as for a landing gear or an individual avionics component),
which would normally be included in what the prime contractor pays
the supplier for the subsystem. Cost-collection systems do not
normally identify such costs separately. However, the study includes
prototype testing if the activities were of sufficient scope and would
have otherwise been included in the EMD phase (now called system
development and demonstration [SDD]).4 On the other hand,
various types of postproduction follow-on testing and production
acceptance testing were not included. We also did not collect data on
_____________
2 These are costs often referred to as direct budget authority, since they are paid for using funds appropri-
ated for T&E in the DoD budget.
3 One of the purposes of the study was to provide more-current cost, technical, and programmatic data
to the cost community. In a few cases, the data we collected were slightly older than our nominal 10
years but were not generally available within the cost organizations and thus would be a useful resource.
4 T&E costs from the demonstration and validation (DEM/VAL) phases of the F-16 and AV-8B were
included in the FSD/EMD totals because of their scope.
advanced concept technology demonstration (ACTD) programs,

although we did discuss them in our interviews.
Because there are expected to be fewer new development pro-
grams in the future, we attempted to include several programs repre-
senting major modifications to existing systems, when enough data
were available. This was also a consideration in selecting parameters
for cost relationships.
Since the purpose of the study was to document the cost of
testing as it is currently being conducted, test procedures or test suffi-
ciency were not assessed.
Organization of the Report

Chapter Two provides a basic overview of T&E for cost analysts. The
topics include:
• the T&E process
• types of T&E
• process flow and documentation
• resources and facilities.
Chapter Three summarizes trends affecting DoD testing and is
based on our interviews with Air Force and Navy test and program
personnel. Chapter Four describes how T&E costs are accounted for
and what the typical data sources are and provides an analysis of cost
trends in DT and OT. Chapter Five recommends approaches for
estimating or assessing the realism of program system T&E costs,
while Chapter Six offers conclusions and other recommendations.
Appendixes A and B consist of brief summaries of the DT and OT of
recent aircraft and guided-weapon programs, respectively. Appendix
C offers an excerpt from the relevant military handbook.
TR-114-AF is a limited-distribution supplement to this report
containing proprietary cost data for the programs described in
Appendixes A and B.
CHAPTER TWO
The T&E Process
T&E involves the collection and analysis of data on the actual or

projected performance of a system or its components. As Chapter
One explained, T&E has three basic functions:
• providing designers with insight on the relationship between
expected and actual performance
• helping managers assess design maturity
• assuring prospective users that the system can perform its
intended mission.
The testing process can be divided into two major categories:
developmental and operational. Developmental testing (DT) is per-
formed at the part, subsystem, or full system level to prove the valid-
ity or reliability of the design, materials used, etc. The results of DT
are used to modify the system design to ensure that it meets the
design parameters and system specifications. Although operational
testing (OT) relies in part on the results of DT, it is designed to test a
system in its operational environment, where operational personnel
(rather than technicians) would be responsible for operating, main-
taining, and repairing the system in a realistic environment.
The intensity and duration of T&E activities vary as the pro-
gram progresses through the acquisition process (see Figure 2.1). In
the concept and technology development (CTD) phase, the T&E
working-level IPT (WIPT) is formed, and an evaluation strategy is
developed to describe the early T&E approach for evaluating various
system concepts against mission requirements. M&S activities also
begin at this time.
13
Figure 2.1
T&E Phases Within the Acquisition Process
Intermittent activities of varying scope

Focus of RAND T&E study
Phase Concept & System Production Support

Technology Development Development
Development & Demonstration
Milestone A B C FRP
System
enhance-
ments
Development
testing
EOA OA IOT&E FOT&E

Operational
testing
FRP = full-rate production.

EOA = early operational assessment.
OA = operational assessment.
RAND MG109-2.1
During the SDD phase, T&E focuses on evaluating alternatives;

components; and eventually, system performance and suitability, as
specified in the mission needs statement (MNS), the operational
requirements document (ORD), the acquisition program baseline,
and the T&E master plan (TEMP).
Once the system enters the production and deployment phase,
full system-level testing predominates. Key developmental, opera-
tional, and live-fire tests of production-representative articles must be
complete before a decision to move beyond low-rate initial produc-
tion (LRIP) can be made. Follow-on OT is conducted to complete
any deferred testing, verify correction of deficiencies uncovered in
previous testing, and refine operational employment doctrine.
T&E continues into the operations and support phase to sup-
port evolutionary or block upgrades. A large program may have a
The T&E Process 15
standing T&E organization, located at a government test center, to

ensure continuity, maximize test program efficiency, and smooth
transitions from upgrade to upgrade.
Congress, in its role of funder of DoD programs, is among the
organizations the T&E process must satisfy and has incorporated its
requirements into law. For example, 10 U.S.C. 2399 specifies that
organizations independent of the developing activity must conduct
realistic OT before a program can proceed beyond low-rate produc-
tion. DOT&E has oversight of this process and must submit annual
reports to Congress.
Starting in 1987, Congress has also required that major systems
and weapon programs undergo realistic survivability and lethality
testing before proceeding beyond low rate production (10 U.S.C.
2366). Later sections in this chapter describe operational and live-fire
testing in greater detail.
Defense acquisition, which includes T&E activities, has contin-
ued to evolve since the early 1990s as increasing pressures on defense
budgets have driven the rationalization and streamlining of the proc-
ess. The current emphasis is on time-phased system requirements so
that improvements in capabilities are added incrementally, as sup-
porting technologies mature. This strategy, often called evolutionary
acquisition or spiral development, presumably reduces costs, risks, and
development time.
Another goal of many recent acquisition reform initiatives has
been increasing the flexibility of the process. While some statutory
and regulatory requirements remain, emphasis on tailoring the proc-
ess to fit individual programs has increased. As an integral part of the
development process, T&E is also affected by these changes. The
effects may range from eliminating duplicative testing or reporting to
allowing the system contractor wide latitude in planning and execut-
ing DT, with the government specifying only the system-level per-
formance the contractor must demonstrate.
The sections below and Chapter Three discuss specific T&E
process initiatives. Since additional changes to the DoD acquisition
process were being drafted as this monograph was being completed,
the DoD acquisition directives should be consulted for the most cur-
rent process information.
Types of Testing
Developmental Test and Evaluation

Developmental T&E (DT&E) is the T&E conducted by the devel-
oping organizations to determine whether the system meets its tech-
nical and performance specifications. The DoD program manager has
overall responsibility for DT&E. With acquisition reform, the system
contractor performs most detailed planning and test execution with
the oversight of the T&E WIPT. The T&E WIPT normally includes
representatives from
• the government program office
• the system contractor
• government test facilities
• the OT activity
• other organizations participating in or supporting testing.
The T&E WIPT is established during the CTD phase or, if
there is no CTD phase, before the program is designated an acquisi-
tion program. One of the WIPT’s first duties is to plan the approach
to be used for evaluating system concepts against the requirements
specified in the ORD or MNS and for the use of M&S. The resulting
evaluation strategy also serves as a basis for the TEMP.
The TEMP is submitted before Milestones B and C for all
acquisition category (ACAT) I programs.1 This document describes
the structure, objectives, activities, schedule, and resources required to
execute the planned test program. It is also updated before OT peri-
ods.
_____________
1 An ACAT I program is one that the Secretary of Defense has designated as a major defense acquisition
program. Such a program normally involves an expenditure of more than $365 million (FY 2000$) for
research, development, test, and evaluation (RDT&E) or more than $2.19 billion (FY 2000$) for pro-
curement (10 USC 2430).
The T&E Process 17
The contractor normally develops an integrated test plan to

document and coordinate the detailed test planning for all DT&E
activities.
T&E activities take place throughout system development.
DT&E supports the system engineering process in evaluating the fea-
sibility and performance of alternative concepts and identifying risk
areas. It supports designers by evaluating component-, subsystem-,
and system-level performance, often highlighting areas for improve-
ment. DT&E also provides managers with objective assessments of
the progress and maturity of the development program. For programs
using a performance-based acquisition approach, DT&E is the pri-
mary means of visibility into contractor progress and system perform-
ance. Both the OT community and users find the data generated
during DT useful for projecting the eventual operational utility of the
system.
A key role of the T&E WIPT is to plan and coordinate the
DT&E requirements of all organizations to maximize the utility of
test data generated and minimize unnecessary conduct or repetition
of tests. One example of this is the elimination of a dedicated period
of government technical evaluation (TECHEVAL) for most systems.
Government DT organizations now use data generated during DT to
support their evaluations. Dedicated government DT is normally
used only for areas of government concern that the contractor does
not or cannot test adequately. Likewise, integrating OT personnel
into a CTF allows what is referred to as combined DT and OT, in
which the tests support both objectives. Thus, OT activities can pro-
vide meaningful feedback earlier in the development process, as well
as reducing or eliminating duplicative testing during the dedicated
operational evaluation (OPEVAL). In general, it is preferable to meet
government oversight requirements through independent evaluation
of jointly collected data.
Another function of DT&E is to demonstrate that the devel-
opment phase is essentially complete and that the system is ready for
dedicated OT. The program manager must certify this readiness.
Finally, successful production qualification testing of low-rate
LRIP articles before the full-rate production decision demonstrates
that a mature production process is in place. DT&E continues at a

lower level throughout the life of the program to evaluate both cor-
rection of deficiencies identified in testing or operation and system
enhancements.
Operational Test and Evaluation

In operational T&E, service test organizations conduct tests in opera-
tionally realistic conditions against threat or threat-representative
forces to determine whether the system is operationally effective and
suitable. Operational effectiveness evaluates the ability of the system to
perform its assigned mission. Operational suitability evaluates the
ability of operational personnel to operate and sustain the system in
peacetime and wartime environments and is a function of such char-
acteristics as reliability, maintainability, availability, supportability,
transportability, compatibility, and adequacy of proposed training
procedures.
The Air Force Operational Test and Evaluation Center
(AFOTEC) and the Navy’s Operational Test and Evaluation Force
(OPTEVFOR) are normally involved in system testing from the early
stages of development. Their role is to provide feedback to developers
on operational issues and to provide periodic operational assessments
based on M&S and DT. Early operational assessments, which are
done before the SDD phase, use studies, M&S, lab trials, demonstra-
tions, and prototypes to evaluate alternatives and the level of risk and
estimate military utility. Operational assessments, done during SDD,
use engineering development models or production-representative
systems. Critical operational issues (COIs) are the questions that OT
must answer. These are derived from ORD requirements and give
focus to OT planning and testing. Measures of effectiveness (MOEs)
are used to evaluate performance on COIs.
Since most programs use what is termed combined DT and OT,
planning for DT also considers OT objectives. The resulting tests can
thus provide data useful for both DT and OT purposes. This mini-
mizes expensive testing, while providing timely data for analysis of
technical performance (DT) and operational performance (OT).
The T&E Process 19
Before any ACAT I or II program can proceed beyond LRIP, it

must meet a statutory requirement for independent testing of produc-
tion or production-representative articles in an operationally realistic
environment to demonstrate the system’s performance for decision-
makers.2 The Air Force refers to this testing as initial operational
T&E (IOT&E), and the Navy refers to it as OPEVAL. The appro-
priate OT agency conducts the tests in operationally realistic scen-
arios involving military personnel trained as users and maintainers.
Statute prohibits the system contractor from participating in IOT&E
in any roles other than those planned for the combat operation of the
system.
Because OT personnel are extensively involved in earlier testing,
dedicated IOT&E is normally much shorter than DT. In effect,
IOT&E can be considered a “final exam” and is therefore a signifi-
cant milestone in the system development. Because of its importance,
the program manager is required to certify the system’s readiness for
dedicated OT.
While OT, by its nature, attempts to maximize realism, full-
scale live testing is not practical in some circumstances. In these cases,
the OT agency, with the concurrence of DOT&E, can use data from
other testing, such as hardware-in-the-loop testing and M&S,3 to
independently evaluate likely operational effectiveness and suitability.
The OT agency’s involvement does not stop with the comple-
tion of dedicated OT. It must forward an IOT&E report to the Sec-
retary of Defense and Congress. The agency also must conduct
follow-on operational T&E (FOT&E) after the full-rate production
decision to
• finish any incomplete IOT&E testing
• verify correction of IOT&E deficiencies
_____________
2 An ACAT II program is one that a service secretary estimates will require an expenditure of $140
million (FY 2000$) in RDT&E funds or more than $660 million (FY 2000$) in procurement but less
than the thresholds for an ACAT I program (see footnote 3) (DoD 5000 Interim Guidance [since super-
seded by new guidance]).
3 Hardware-in-the-loop testing involves exercising actual systems or subsystems in a controlled envi-
ronment using simulated inputs.
• refine estimates or projections from IOT&E

• evaluate significant changes to the system design or employment
• evaluate new operational concepts or environment.
Additional OT is conducted whenever system modifications
materially change that system’s performance (DoD 5000.2-R).
FOT&E usually involves the using command. DOT&E or the OT
agency determines the quantity of test articles for dedicated OT.
Multiservice Operational Test and Evaluation

Systems that multiple services will acquire or use must undergo
multiservice operational T&E (MOT&E). The designated lead ser-
vice has primary responsibility for the test program and test proce-
dures, with participation from the other services. A service with
unique requirements does its own planning, testing, and funding for
them. Because of differences in employment, test results that may be
satisfactory for one service may not be for another.
Qualification Operational Test and Evaluation

Qualification operational T&E (QOT&E) is an Air Force term for
the testing of modifications or new employment of a system for
which there is no RDT&E-funded development.
Live-Fire Test and Evaluation

Although live-fire testing of manned systems for survivability (a com-
bination of susceptibility and vulnerability) and missile and muni-
tions for lethality had previously been a part of system testing, the
1987 Defense Authorization Act mandated that all “covered systems”
undergo realistic full-scale live-fire testing before proceeding beyond
LRIP. DoD 5000.2-R defines covered system as
• a major system within the meaning of that term in 10
U.S.C. 2302(5) that is,
– user-occupied and designed to provide some degree of
protection to its occupants, or
– a conventional munitions program or missile pro-
gram; or
The T&E Process 21
• a conventional munitions program for which more than

1,000,000 rounds are planned to be acquired; or
• a modification to a covered system that is likely to affect
significantly the survivability or lethality of such a sys-
tem.
Under additional legislation in 1994, Congress directed

DOT&E oversight of live-fire T&E (LFT&E). DOT&E may grant
waivers when realistic system-level LFT&E would be “unreasonably
expensive and impractical.” DOT&E must approve waivers of full-
scale live-fire testing in combat configuration before Milestone B or
program initiation. The waiver request must describe the proposed
alternative LFT&E strategy.
LFT&E involves testing at the component, subsystem, and sys-
tem levels. It is normally part of DT to allow identification and
implementation of any necessary design changes before full-rate pro-
duction. Actual live-fire tests or shots are normally preceded by exten-
sive modeling of damage or lethality to determine shot lines and
sequence of shots to maximize the value of data collected and to assess
model validity.
Test and Evaluation Process Steps

All effective T&E is tailored to the system performance requirements,
the nature and maturity of the system under test, and the time and
resources available. However, the complexity of testing a major
weapon system and the coordination required dictate using a general
process or framework as a guide for effective planning and execution
of the test program.
Figure 2.2 illustrates the interrelated aspects of the T&E proc-
ess. This cube relates the test processes, T&E resources, and system
maturity and illustrates the cyclical and concurrent nature of T&E.
The steps in the test process are repeated as the system matures. The
objective is to identify deficiencies as early as possible to avoid the
cost and delays inherent in redesign and retest of a more complete,
Figure 2.2
System Maturity, Test Processes, and Resources
Report results
s
e ss Post-test analysis n ge
oc ra
s
pr
t
ie
Conduct test en
lit
st
ci
Te nm
fa
Pre-test analysis ro
st
vi
te
n
-e
em
Determine objectives n s
pe ie
st
lit
bs
sy
O i
tie n la
ac
d
Fielded system (P3I/mod)
f
le
ac io
L
al
s
IT
t
st
t f ra
H
Integrated system
In
ili
en eg
n
tio
m nt
Component
i
ul
Subsystem
su tem
Concept
System
sim
s
re
Sy
s
d
ce
an
ur
ea
g
so
M
e lin
re
od
E
T&
M
System maturity
RAND MG109-2.2
and therefore more complex, article. Each step of this generic T&E
process is described below.4
Determine Test Objectives

Specific test objectives are normally derived from the system-level
requirements stated in the MNS, the ORD, and the TEMP. These
flow down to lower levels through subsystem and component
requirements that the contractor derives to support design and testing
at each level. Given these performance requirements, test planners
specify the article to be tested and the test conditions, methodology,
and data required by developing a list of specific test events. An
example of a test event would be measurement of key parameters at a
specified altitude, speed, attitude, and configuration. The progression
of these events is called the test matrix. The objective of the events in
the test matrix is to gain sufficient knowledge about the performance
_____________
4 Air Force Manual 99-110 (U.S. Air Force, 1995) discusses the generic process in more detail.
The T&E Process 23
of the test article under a range of conditions to reduce the risk of

proceeding to the next step in development to an acceptable level.
Conduct Pretest Analysis

Before testing begins, the test matrix is evaluated to ensure that it will
produce the required data while making efficient use of test resources.
Current knowledge, simulations, and other analyses are used to help
predict the results to be observed. The responsible test organization
develops a detailed test plan, making allowances for unplanned
delays.5 Outside technical experts then review the detailed test plan.
Conduct Test
Prior to conducting the test, the team must select specific test points
from the test matrix. In some cases, this may not be a simple progres-
sion from basic to more challenging tests because of such external fac-
tors as concurrent testing, availability of a properly configured test
article, suitable environmental conditions, and required supporting
assets. In addition, opportunities for achieving multiple test points
during a test event require careful planning and coordination to
maximize the productivity of test resources. Efficient and timely exe-
cution of a test program becomes a complex balancing act requiring
coordination (and often flexibility) among all participants.
Perform Post-Test Analysis

Modern test programs produce large volumes of data, which must be
reduced and converted to a useful form for evaluation. Most test
activities can produce rapid, “quick-look” reports to give some indica-
tion of test outcomes before conducting more-detailed analysis of the
test data. Quick-look results can be used during the test itself to verify
that suitable data were collected and that the next test can proceed.
After the test, the data are analyzed extensively and compared to pre-
dicted values. If the predicted values do not match those observed,
further analysis is required to determine whether the predictions, test
_____________
5 Many test organizations use a test efficiency planning factor of 80 percent to allow for weather,
equipment problems, additional testing, or nonavailability of the test article.
conduct, or system under test is responsible. In many cases, a retest

must be conducted to validate corrective action.
Report Results
Test results are reported in a variety of ways, depending on the sig-
nificance of the test and the intended recipients. At the end of DT
and OT, formal reports are submitted and retained to provide a
record of test program execution and results.
If a valid trial reveals a discrepancy between predicted results
and actual system performance, the prediction algorithms are reana-
lyzed and adjusted as necessary. If a test article deficiency is found, its
cause(s) must be isolated, which may require further testing. Once
the cause(s) is (are) identified, a design change may be made.
As the type of article being tested progresses from component to
subsystem to system, design changes can become increasingly com-
plex and costly because they can affect the operation of other compo-
nents or performance in other operating regimes. As a result,
extensive retests (regression testing) may be required to ensure per-
formance results from previous tests have not been compromised.
This situation leads the program manager to a dilemma: Is it better to
reduce risk through extensive component and subsystem testing, with
known increases in cost, or to save costs through reduced testing,
with the potential for much greater disruption and cost if perform-
ance deficiencies are discovered later in development. Obviously, the
optimal balance between testing and risk depends on the maturity
and complexity of the system and on the criticality of potential per-
formance shortfalls.
Selected Test-Related Documentation

This section briefly describes typical documentation that may be use-
ful to an analyst attempting to assess the cost and scope of T&E for a
major acquisition program. A number of other reports may provide
specific data needed for a thorough analysis, but the reports described
below generally provide most information required.
The T&E Process 25
Analysis of Alternatives
The analysis of alternatives (AoA) is a comparison of alternative
approaches for meeting the mission needs described in the MNS (or
in some cases, the capstone requirements document). AoAs are per-
formed by representatives of the user community, although represen-
tatives of the program manager may participate. An AoA is required
before Milestone B (or Milestone C, if there is no Milestone B). For
the test community, the AoA provides COIs and MOEs to be used in
OT.
Operational Requirements Document

The ORD translates the general military requirements contained in
the MNS (or capstone requirements document) into specific user and
affordability requirements. It also describes how the system will be
employed and supported.
To support evolutionary acquisition, ORDs should contain
time-phased capabilities. To support the CAIV process, the docu-
ment should provide threshold and objective values for key perform-
ance parameters, to enable trade-offs during system development.
In addition, the ORD requirements should be prioritized, to
guide system developers as they make affordability, schedule, and
risk-reduction trade-offs. Early versions of ORDs do not necessarily
fully define capabilities of subsequent spirals or blocks. Thus, actual
operating experience can help refine requirements for subsequent
iterations.
The initial ORD is prepared for Milestone B or program initia-
tion. Periodic updates add new details as the development program
matures. Because the ORD represents the approved objectives of the
users of the system, it is particularly useful for developing OT plans.
Test representatives from the services should participate in ORD
development to ensure that the requirements are testable.
Test and Evaluation Master Plan

The TEMP provides the overall structure for all parts of the test pro-
gram and provides the framework for developing detailed test plans.
The TEMP describes the development, operational, and live-fire
testing; provides a test schedule; and outlines the required resources,

including the funding, test articles, test sites and equipment, threat
representation, targets, manpower, training, and operational force
support. The plan also presents COIs, MOEs, critical technical
parameters, and suitability thresholds and objectives derived from the
ORD. If there are service-unique test requirements for a joint pro-
gram, they appear in an appendix.
The TEMP is preceded by an evaluation strategy submitted
within six months after Milestone A or program initiation. The
TEMP itself is prepared by the program manager and the T&E
WIPT, with input from the OT agency on operational testing. The
TEMP is first submitted for Milestone B, then is updated before each
OT period and for Milestone C, the full-rate-production decision
review, and any significant program changes.
Test Resources
The T&E process requires a variety of resources, including contractor
and government test facilities, test ranges, manpower, training, flying
time, support equipment, threat systems, M&S, instrumentation,
communications, range equipment and facilities, data protection and
security systems, munitions, and targets. While many of these needs
are relatively straightforward to estimate for a given test, the timely
availability of appropriate infrastructure to support testing is critical
to the success of any test program. Test facilities may be required
from any of the following general categories:
• M&S
• measurement facilities
• system integration laboratories
• hardware-in-the-loop facilities
• installed system test facilities (ISTFs)
• open-air ranges.
The T&E Process 27
Figure 2.3
Number of Trials by Various Test Methodologies
Number of trials
Digital modeling
measurement
Precision
Int. lab.
testing
OAR
HITL
Modeling and simulation ISTF
System maturity
RAND MG109-2.3
These categories can be thought of as representing a spectrum

based on level of effort per trial as shown in Figure 2.3. In the figure,
the level of effort per trial increases from left to right, as the number
of trials decreases. Each is described below.
Digital computer models are used throughout all phases of system
development to simulate and evaluate components, subsystems, and
systems. The use of M&S for testing parallels its use in the develop-
ment process, progressing from relatively generic high-level modeling
for evaluating system concepts and alternative architectures to
increasingly detailed models as the system design matures and as key
system parameters are determined. Typically, as the models mature,
they increasingly rely on physical or engineering relationships that
can be validated empirically through observation or testing. These
component model results can often be aggregated, or the models can
be integrated, to produce a system-level model that will predict, with
considerable accuracy, the events and conditions that the model

architecture and data fully describe. Once validated, these models can
be used to predict system behavior under a wide range of conditions
at considerable savings in test time and cost.
Digital models consist of algorithms that can be run by provid-
ing fixed input parameters, by having the model develop and modify
scenarios dynamically within the simulation, or by using an operator
in the loop to capture the interaction of human operators with the
system in a virtual environment. However, despite experience with
the use of validated models, it is unlikely that they will ever fully sup-
plant live testing. Even in areas that are well understood, it is not
unusual for live testing to uncover problems that were not apparent
in the simulations. Examples include wing-drop effects in the
F-18E/F, buffeting of stores in certain flight conditions on a number
of programs, and target penetration prediction errors in the Joint Air-
to-Surface Standoff Missile (JASSM). These problems seem to occur
most frequently when complex interactions are either poorly under-
stood or inadequately modeled.
Models must be verified and validated before use. Verification
involves analysis and testing to ensure that the model functions as
designed. Validation establishes that the model, when used over some
range of conditions, acceptably represents reality. Accreditation is a
further step, certifying that the model is appropriate for specific pur-
poses. The development, verification, validation, and accreditation of
models is a significant part of early system development and must be
planned and budgeted so that it is in place to support T&E.
The benefits of M&S are that they can
• simulate a large number of trials that differ in predictable ways
over a short period
• substitute for live testing when safety, availability of appropriate
assets or test conditions, environmental factors, or expense make
live testing impractical
• support trials in which selected parameters are varied in con-
trolled ways, while others are held constant
• identify system shortcomings before live testing
The T&E Process 29
• identify or refine issues and conditions for live tests to maximize

the value of data collected
• increase confidence that the selected live-test data can be inter-
polated or extrapolated.
Measurement Facilities
These facilities accommodate accurate measurement of parameters of
interest under controlled conditions. Such facilities are normally gen-
eral purpose, although some modification or special equipment may
be required for a specific test program. The following are examples of
measurement facilities for aircraft and missiles:
• wind tunnels
• propulsion test facilities
• signature-measurement facilities
• environmental measurement facilities
• warhead test facilities.
System Integration Laboratories

These laboratories support integration of hardware and software
components in a controlled environment. Components under test
can be bench-tested with other simulated components and software.
These are generally contractor facilities, although the government
may also have such laboratories for testing purposes.
Hardware-in-the-Loop Facilities
Hardware-in-the-loop testing involves exercising actual system or
subsystem hardware in a controlled environment using simulated
inputs. Hardware-in-the-loop facilities can simulate the systems and
threats with which the test article must interact. They are most often
part of contractor facilities, although the government maintains them
as well.
Installed System Test Facilities

ISTFs allow ground testing of complete systems while they are
installed physically or virtually on the host platform. Test facilities
may be electronically linked to test data interchange. For instance, an
aircraft in a hangar may use its weapon control system to provide

commands to an instrumented missile located in another facility. In
this way, hardware-in-the-loop facilities may become virtual ISTFs.
Examples include anechoic chambers and structural load test facili-
ties.
Open-Air Ranges
Open-air ranges are used for flight testing aircraft and for a variety of
guided-weapon tests, from captive carry through live fire. These are
the most resource-intensive test facilities, and nearly all are govern-
ment owned. Table 2.1 lists the primary Air Force and Navy aircraft
and missile open air ranges.
Table 2.1
Primary Air Force and Navy Open Air Ranges for Aircraft and Missile Testing
Range Location
Air Force Flight Test Center Edwards AFB, California
Air Force 46th Test Wing Eglin AFB, Florida
Holloman AFB, New Mexico
Air Force Air Warfare Center Nellis AFB, Nevada
Air Force Utah Test and Training Range Hill AFB, Utah
Naval Air Warfare Center—Aircraft Division Patuxent River, Maryland
Naval Air Warfare Center—Weapons Division China Lake, California
Pt. Mugu, California
CHAPTER THREE
Trends in Test and Evaluation
A key objective of this study was to identify changes in the practice of

T&E and, to the extent possible, identify their likely effects on the
cost of T&E for future aircraft, missiles, and guided munitions. An
initial step was to compile a list of those trends we judged most likely
to affect T&E cost now and in the immediate future. Offering the
results of this process as topics for discussion in advance of our inter-
views with program and test facility personnel provided some useful
structure. We compiled this list from various references, with feed-
back from Air Force, Navy, and OSD personnel. The rest of this
chapter addresses each of the trends we examined.
The trends examined were
• acquisition reform
• M&S
• testing of software-intensive systems
• combined and multiservice testing
• contractor versus government test facilities
• live-fire testing
• warranties.
Acquisition Reform
In defense procurement, acquisition reform generally refers to a broad
emphasis on eliminating activities that do not add value and ensuring
that the remaining activities are as cost effective as possible. A previ-
31
ous RAND report (Lorell, 2001) addresses the general subject of the
effects of acquisition reform on cost estimating. In that report, the
authors provide a taxonomy of acquisition reform measures:
• reducing regulatory and oversight compliance costs
• adopting commercial-like program structures and management
• using multiyear procurement.
In T&E, all acquisition reform initiatives fall into the first two
categories, reducing compliance costs and commercial-like practices.
Using the taxonomy described in Lorell (2001), the acquisition
reform principles that specifically apply to T&E are
• requirements reform (ORD flexibility)
• contractor design flexibility and configuration control (TSPR)
• commercial insertion (COTS/NDI).
In addition, the ACTD approach, which uses mature technolo-
gies to demonstrate new operational capabilities in abbreviated dem-
onstrations, often manifests all these principles. 1 ACTDs are not con-
sidered acquisition programs.
The following subsections discuss each of these approaches to
acquisition reform in T&E.
ORD Flexibility
The ORD represents the set of mission requirements a new weapon
system must meet. It describes the performance that the new system
is expected to provide to the operating forces and, throughout the
development process, is the authoritative guide for the functional
capability the system must attain. The ORD describes these
capabilities in terms of key performance parameters, which both pro-
vide guidance for the design team and are the basis for the test pro-
gram.
The ORD development process has, in the past, been criticized
as being a “wish list,” derived primarily from user desires with rela-
_____________
1 OT personnel often conduct or observe ACTD and attempt to provide an assessment of potential mil-
itary utility.
Trends in Test and Evaluation 33
tively little consideration of the cost of achieving specified levels of

performance. One of the fundamental tenets of CAIV is that user
“requirements” should allow maximum flexibility, so that degrees of
performance can be traded against cost to arrive at a “best value” sys-
tem, one that recognizes that key performance parameters are not all
equally important and that incremental levels of performance were
not equally valuable to users. 2
The CAIV approach involves setting true minimum perform-
ance requirements as thresholds that must be met, and providing
incentives for reaching or exceeding the higher, objective or target
requirements within the overall cost objective. In this way, the con-
tractors, working with the government IPTs, can attempt to maxi-
mize value for the government acquisition dollar. This also provides
the test community more definitive guidance for assessing operational
utility. Most ORDs have supportability and maintainability targets to
reduce the tendency to save acquisition cost at the expense of long-
term operations and maintenance.
In addition to flexibility in program requirements, acquisition
reform also encourages simplification of the ORD to a minimum
number of top-level performance measures in an effort to reduce or
eliminate the perceived tendency to overspecify. This is intended to
promote program flexibility and avoid well-intentioned but ineffi-
cient constraints on candidate approaches and designs.
The ORD, in addition to providing guidance for the develop-
ment of derived specifications for the developers, is also the source of
requirements for the test program. Key performance parameters are
broken down into critical issues, which are the questions T&E must
answer. Critical issues must be translated into testable characteristics,
often called MOEs, against which specific measures of performance
may be observed during testing.
Since the ORD provides the performance requirements for sub-
sequent T&E, it is important to get effective input from the T&E
community while the ORD is being drafted. This is to ensure that
_____________
2 In contrast, in the earlier design-to-cost approach, the developer’s flexibility was limited by a set of
requirements that was, for all practical purposes, fixed.
the performance thresholds and objectives are clear and can be veri-
fied by testing.
A good example of the importance of the ORD in T&E was
provided when DOT&E judged that the Predator unmanned aerial
vehicle was not operationally effective or suitable, although it was
deployed in Afghanistan and operated with notable success there. The
DOT&E evaluation was based on the fact that the system had to be
operated well below the requirements set forth in its ORD, the stan-
dard against which the tests had to be conducted. If the ORD lacks
“crawl, walk, run” performance requirements, system developers and
testers must use ultimate performance thresholds as their guidance,
even if lower levels of performance would provide useful capability.
These phased requirements are expected to become the norm with
the DoD emphasis on evolutionary acquisition.
Total System Performance Responsibility

The TSPR concept is an attempt to emulate the commercial research
and development environment, in which the company has nearly
complete control over (as well as complete responsibility for) the
design, development, and testing of a new product. While user
requirements, preferences, and budget constraints are significant
inputs, the company itself decides on development approaches, spe-
cific designs, budget allocations, risk mitigation, and any necessary
infrastructure investments. Balancing this autonomy is direct feed-
back from the market. In the defense environment, the discipline of
the market must be created by careful specification and communica-
tion of requirements and well-crafted contractual incentives that
reward success and penalize poor performance. The presumption is
that the possibility of increased profitability will provide incentives
for the contractor to be innovative in design and vigilant in reducing
unnecessary expense.
TSPR has significant implications for the conduct of T&E.
First, it means that the contractor will have primary responsibility for
designing and conducting the DT program. While still being
required to demonstrate compliance with contract specifications, the
contractor has, in many cases, primary responsibility for determining
the timing and nature of testing. This may include determining the
test facilities to be used, the number of test articles, and the amount
of testing. Of course, the test IPT normally contributes to these deci-
sions, since the government must ultimately evaluate the results to
determine specification compliance. However, if government
representatives request additional tests for risk mitigation or evalua-
tion in a particular facility and if the contractor disagrees, the gov-
ernment may have to provide additional funding through a contract
modification.
Second, unless the contract specifies otherwise, TSPR may mean
that certain contractor-developed data will not be available to the
government test community. For example, the contractor may model
airflow as part of the design process. The resulting data might be use-
ful to the government for simulating other events, such as weapon
separation. But if certification of the weapons in question are not a
part of the original development program, the government may have
to recreate the data through its own testing or may have to purchase
the data from the contractor separately.
Third, representatives from several programs noted that contrac-
tors having configuration control occasionally made modifications
that affected interfaces with other systems. Thus, government per-
sonnel needed to track contractor-initiated changes closely for effects
on related systems.
Finally, TSPR highlights the importance of close cooperation
and frequent interaction with government test experts, especially the
OT community, so that contractor personnel understand user
requirements and operating procedures. Both contractors and gov-
ernment program personnel noted the importance of early identifica-
tion of potential problems. This is particularly important when
documented requirements or specifications have been reduced to fos-
ter contractor innovation. Authoritative government input early in
the process can save costly redesign and prevent potential contract
disputes. Another benefit of frequent interaction is that it improves
the familiarity of government personnel with the system, preparing
them for evaluating test data.
Commercial Insertions (COTS/NDI)

Although full COTS systems are rare in DoD aircraft and guided-
weapon programs, the use of COTS components, when feasible, is
becoming increasingly common, both because such components are
less expensive than custom military items would be and to take
advantage of advances in commercial technologies. A related trend is
the use of an existing military system, or NDI, that the U.S. DoD or
another country has developed for other applications. Both COTS
and NDI have perceived advantages:
• existing production base
• reduced or nonexistent development costs
• low technological risk
• shorter lead times than for new development.
From a testing perspective, we found that using COTS and
NDI components does not, in general, significantly reduce system-
level testing because the military environment is often quite different
from their typical commercial employment. In many cases, normal
system-level tests were completed without failures, and COTS com-
ponents were incorporated into the final design at significant cost
savings. In other cases, particularly when entire systems were COTS
or NDI, what were assumed to be minor modifications to accommo-
date DoD requirements became major redesigns, in some cases
growing into significant development and test programs.
As part of this study, we interviewed personnel from three air-
craft programs that could be considered predominantly NDI: T-45,
Joint Primary Aircraft Training System (JPATS), and C-130J. In the
T-45 and JPATS cases, the acquisition strategy involved buying U.S.-
built versions of existing foreign aircraft with modifications to adapt
them to U.S. requirements. In the case of the T-45, the existing Brit-
ish Aerospace (BAe) Hawk was chosen to replace the current Navy
intermediate and advanced strike trainers. A firm-fixed-price contract
was awarded to McDonnell Douglas to modify the Hawk for aircraft
carrier operations. Because it was an existing airframe, little M&S was
planned. Initial OT identified major problems in aircraft handling,
both in flight and on the ground. Correcting these deficiencies
required much more redesign and iterative testing than anticipated.

Because neither the contractor nor the government had planned for
extensive rework and testing, the program’s schedule slipped several
times. The aircraft finally completed OT in 1994, ten years after the
program entered FSD.
JPATS was the first aircraft to be designated as an acquisition
streamlining pilot program. Streamlining acquisition procedures and
limiting competition to variants of existing aircraft were expected to
reduce the time needed to acquire and field a replacement for Air
Force and Navy primary flight trainers. The complete JPATS
includes aircraft simulators, training devices, courseware, a training
management system, and contractor logistics support, but our discus-
sions with the program office were limited to the aircraft portion of
the system.
Source selection included flight evaluation of seven competing
aircraft. The Beech (now Raytheon) candidate was selected, and the
aircraft reached Milestone II in August 1995. The first two phases of
qualification T&E (QT&E) were performed on a prototype aircraft
because a production-representative aircraft was not available. One of
the acquisition-reform initiatives for JPATS involved relying on Fed-
eral Aviation Administration (FAA) certification for portions of the
flight-test program. It became apparent, however, that FAA certifica-
tion is a cooperative effort between the FAA and the requesting con-
tractor to promote basic airworthiness, not to demonstrate compli-
ance with stringent performance specifications. As a result, the
planned DoD portion of the flight-test program grew from approxi-
mately 50 to 90 percent.
Radio testing is one example of the problems encountered. The
FAA commercial standard for radio performance is a ground check,
which missed significant dead zones in the antenna pattern. Military
testing was required to identify and resolve the problem. Similarly,
spin and recovery testing had to be expanded when FAA standards
were judged insufficient for training military pilots. The contractor,
whose experience with military aircraft testing was limited, did not
foresee the additional testing required and the shortfalls it uncovered.
Most testing was performed at the contractor’s site. As in most TSPR
programs, additional government-directed testing required a contract

modification.
The C-130J is a contractor-initiated update to the C-130 series
of medium-range transport aircraft. Although it retains the exterior
and interior dimensions of its predecessors, more than 70 percent of
this model is unique, including integrated digital avionics, a redes-
igned flight station, a new propulsion system, and cargo compart-
ment enhancements (DOT&E, 2000). Allied countries have pur-
chased C-130J variants, and the U.S. Air Force is now procuring it
under a commercial acquisition strategy.
The initial T&E approach was to supplement FAA certification
with government QT&E in specific military areas of interest. After
some delay, a commercial variant that was a conceptual combination
of two production aircraft achieved FAA certification. The contractor
subsequently decided not to maintain FAA certification. The first
seven aircraft had to be modified to bring them into compliance with
their model specification, and software problems have required more
intensive testing than planned. The contractor planned and executed
the DT program, with government DT and OT personnel observing.
The program has instituted a CTF to improve coordination of
QT&E and OT&E data gathering. The test program is phased to
coincide with software revisions. The Air Force conducts its own
limited functional software testing, generally on the aircraft.
The primary issues for the C-130J have been numerous software
problems and the vulnerabilities demonstrated in live-fire testing. As
with T-45 and JPATS, what was intended as a low-risk modification
to an existing system under a commercial-type procurement has
required far more development time and effort than originally antici-
pated.
According to the program office, contractor T&E cost data were
not available to them because of the nature of the contractor-initiated
development effort. As a result, our cost analysis could not include
the C-130J.
In summary, these commercial-type procurements yielded the
following lessons:
• Both the government and the contractor should clearly under-

stand the differences between the performance capabilities of the
system as it exists and those expected in a U.S. military envi-
ronment. The system requirements review is an important
opportunity to define an acceptable system specification clearly.
• FAA certification alone is not sufficient to demonstrate
achievement of most military aircraft performance specifications.
• “Best commercial practices” are not an effectively codified set of
principles like those of common law or accounting. Because
such practices tend to depend on the situation and to be incon-
sistent from contractor to contractor, they may be inadequate
for responsible acceptance of military systems. To be successful,
this approach requires extensive communication and, eventually,
a mutual understanding between the contractor and customer of
specifically what constitutes an acceptable demonstration of sys-
tem performance.
• Even well-proven commercial products should be tested in a
representative military environment. In many cases, they will be
found to be as suitable as they are in their civilian applications.
In some cases, however, the unique demands of military opera-
tions will cause the products to fail, often because of conditions
not typically encountered in civilian operation.
• Because of the lack of leverage with the contractor when the
military market is a small portion of the contractor’s business
base, it is critically important to specify performance and test
requirements carefully for prospective contractors, to put con-
tractual incentives in place to encourage success, and to ensure
enough government insight to recognize impending failure.

M&S generally refers to the use of computer models to emulate a sys-
tem to provide insight into its operation without actually operating it.
This approach is now used extensively throughout the acquisition
process, from evaluating system concepts through operational train-

ing.
In T&E, M&S can be used to identify and quantify key system
parameters that will become key performance parameters, MOEs,
and measures of performance; to model system and component
operation; and to evaluate system and component performance.
M&S is particularly useful when live testing is not practical because
of range, safety, or threat limitations.
M&S is often cited as a transformational advance in T&E tech-
nology. The claims that M&S reduces the requirement for live test-
ing, therefore also reducing the expense of the test program, is par-
ticularly relevant for our study. Because of the importance of this
issue, we specifically highlighted M&S and its potential for cost sav-
ings in interviews with test program managers, test activity personnel,
and cost analysts.
A wide variety of models and simulations are used in various
phases of T&E. Top-down models may use aggregate characteristic
data to represent operation of a system in a wide variety of scenarios,
including force-level engagements. Engineering or physics-based
models are typically based on detailed data and can be used to model
components and subsystems or can be integrated to model an entire
system. Both types can include interfaces with actual components
(hardware-in-the-loop) or human operators (man-in-the-loop). Engi-
neering (or physics-based) models are common in T&E because they
often arise from the engineering design activities of the program, or
the test activities specially develop them. Their data and relationships
are carefully developed and validated, normally starting with a basic
capability and expanding over time. Although these models are based
on “hard” physical or engineering relationships, they may fall short if
the model does not include all relevant parameters, effects, or interac-
tions. In some cases, such elements may be overlooked because they
occur infrequently, because their effects are subtle, or because the
phenomena to be modeled are not completely understood.
Top-level models, on the other hand, are often used to compare
alternatives and set performance thresholds. They are used less often
in T&E applications because of the difficulty of objectively verifying

their fidelity across the full range of interest.
M&S can take a variety of forms in T&E applications, including
• detailed component testing
• subsystem integration
• system simulation and stimulation
• campaign or force-level (system of systems).
The following are some common advantages of applying M&S
in T&E:
• It can overcome or reduce test limitations that are due to range
restrictions, safety, or threat interactions.
• It makes it possible to conduct many trials efficiently that
involve varying known parameters in predictable ways.
• It provides a means of adjusting or testing parameters for
insight, sensitivity analysis, or optimization.
• Once developed and validated, models are extremely valuable for
system upgrades or follow-on development efforts.
The benefits can also carry over to other related test programs (e.g.,
threat simulators) if the models are serviceable and well documented.
Using M&S for testing also has limitations that should be rec-
ognized:
• Level of Fidelity: Although it may sound obvious, the available
models are not always able to reflect real-world outcomes consis-
tently. This generally happens in areas for which the physical
phenomena have not been characterized well or in which there
are complex interactions that are difficult to predict accurately.
Another, more common, problem is limited availability of actual
detailed data for model development and calibration. This type
of data typically requires specialized instrumentation and
repeated tests to collect. Because of the time and expense
involved, the collection of adequate data for non–program-
specific, high-fidelity modeling seems to be fairly uncommon. A
third component of high-fidelity modeling is validating model
operation with observations of actual performance. As with data

collection, this requires conducting physical tests with sufficient
instrumentation to validate model predictions under similar
conditions.
• Range of Applicability: Models are typically developed and
tested over an expected range of input and output values. When
models are used outside these ranges, their outputs may become
unreliable. The current environment, in which program manag-
ers fund M&S development, does not encourage the design or
validation of models for conditions outside the program’s
immediate requirements.
• Cost of Model Development: When “affordability” is a major
criterion, every significant item of cost in a program budget is
scrutinized. In addition to putting the obvious constraints on
the total program, program managers must also carefully phase
activities to match funding available for a given period. Budget
cuts or unexpected problems may require deferring planned
activities to later fiscal years. Because of the time it takes to
develop and test complex software, effective M&S investments
must be made early in the program, well before their intended
use. Unfortunately, this is also a time when most programs are
not well staffed or funded. Thus, without a strong advocacy,
M&S tends to be developed on a limited budget and a con-
strained schedule. These early limitations may manifest them-
selves in ways both obvious (schedule slips) and subtle (limits on
design flexibility and future expandability).
• Integration with Other Models and Hardware: Another
often-underestimated aspect of model development is the
integration of component- and subsystem-level models into
higher-level models. As with individual model development,
early funding or schedule constraints may hinder planning for
integration of various models. Although the benefits of
coordinating the development of lower-level models so that they
can easily become part of an effective and flexible system-level
model are well recognized, budget, schedule, and occasionally
organizational constraints again make this challenging. When a
higher-level model is needed and detailed models are not

available or sufficiently mature to use as components, a top-
down model may be substituted. These models have often been
criticized as being less robust because many of the inputs, and
sometimes the algorithms themselves, tend to be more
conceptual and are rarely subjected to rigorous validation
through actual testing.
In summary, M&S has become integral to most development
programs. It supports testing of components and systems when live
testing is not economically or practically feasible. It often allows
alternative designs to be tested and iterated before commitment
makes changes prohibitively expensive. Comprehensive M&S pro-
grams may also contribute to other areas, such as development of tac-
tics, training, and future system enhancements. There is general
agreement that M&S has reduced the amount of live testing that oth-
erwise would have been required, although other test activities tend to
mask these savings. It also had the affect of making most live testing
more focused and robust than in the past. The Defense Science
Board, in its 1999 review of T&E, found that
Claims of substantial program cost savings attributable to the
increased use of M&S, with a concomitant reduction in testing,
cannot be verified. ... a White Paper prepared by the AIAA
[American Institute of Aeronautics and Astronautics] Flight Test
Technical Committee (FTTC) in late 1988 entitled, “Seeking
the Proper Balance between Simulation and Flight Test,” states
“the members of the FFTC are unaware of any study that has
supported the claim of substantial program cost savings realized
by a significant expansion of the use of M&S with a concomi-
tant reduction in testing.”
Similarly, the Air Force Scientific Advisory Board (1998) concluded

that
Evolutionary improvements in the use of M&S enhance T&E
and may eventually reduce the costs; however, they cannot be
expected to become a cost-effective or technically sufficient
replacement for most physical testing of air vehicles in the fore-

seeable future.
As in system design, M&S has increased the efficiency of system

testing. In such areas as aerodynamic modeling, it has reduced the
wind-tunnel hours required. The Air Force SEEK EAGLE program
can now do computerized fit checks for stores compatibility. In fact, a
number of test personnel admitted to us that the M&S capabilities
for typical programs are being used to such an extent that they are
hard pressed to keep up with the demands for capability and fidelity.
The goal of many in the test community is for M&S to be the
primary iterative test method, with live tests used for validation.
Today, M&S is routinely run in predictive mode before live tests and
is modified as required to reflect observed outcomes.
Testing of Software-Intensive Systems

Most current aircraft and guided-weapon programs can be classified
as “software intensive.” For a variety of reasons, much of the func-
tionality of modern weapon systems is actually implemented in soft-
ware. This has advantages for upgradeability and evolutionary devel-
opment, as well as for improved availability and supportability of
common hardware components. However, developing and testing
complex software presents some of the most difficult challenges in
system development, particularly when a large number of functions
and a high level of integration with other hardware and software
strain the ability of current development processes and tools.
Methods of improving software testing are being investigated
and implemented across a wide range of government and commercial
programs. Although it is beyond the scope of this report to address
this issue in detail, we can summarize some of the approaches relevant
to aircraft and guided-weapon software testing.
A common predictor of future software development problems
is incomplete, unclear, or changing requirements. Although this
problem is well recognized, many programs continue to suffer from
poor definition of software requirements. In addition to leading to
faulty software, poor requirement definition compromises test activi-

ties by making it more difficult to design and execute effective soft-
ware testing; by leading to late discovery of problems and delays in
testing while code is fixed; and by requiring additional time and
effort for regression testing.
Virtually all modern software development involves some form
of structured development approach. For large or complex systems,
this generally involves some form of evolutionary or “spiral” devel-
opment, which introduces additional software functionality in
sequential releases or “blocks.” These control the amount of new code
and functionality that must be tested in each release, in theory
building each subsequent increment on relatively stable, well-tested
modules. While this approach may produce higher-quality software,
several programs noted that it complicates system-level test planning
by restricting testing to the capabilities implemented in the current
release and by possibly increasing the amount of regression testing
required.
In missile programs, the use of “captive carry” flights to test mis-
sile components on an aircraft simulating the missile’s flight is well
established. Similarly, on several larger aircraft programs, flying test
beds were used for testing hardware, software, and the integration of
the two. In general, the programs that used flying test beds judged
them to be very effective. They allowed early DT in conditions
closely simulating the operational environment, without the problems
and expense of testing immature systems and software on the actual
developmental aircraft. For programs with significant software but
without a flying test bed, one or several robust ground-based integra-
tion laboratories were a virtual necessity.
Since testing accounts for a significant portion of the software
development effort, software can be designed for improved testability
by considering and accommodating test requirements throughout the
development process. Since manual test design is labor intensive and
error prone, automated generation of test cases and automated, adap-
tive test algorithms can both improve the quality of testing and
reduce the time required.
In summary, the challenges software-intensive systems present

will continue to grow with the introduction of more complex systems
and with the more-stringent interoperability requirements of the “sys-
tem of systems” approach to defense acquisition. With this increase in
complexity, costs of testing software-intensive systems can be
expected to grow unless testing efficiency can somehow be improved
to offset the complexity increases. The common goal of all the testing
approaches described above is to identify and correct problems as
early as possible, minimizing those discovered in flight testing.
Combined and Multiservice Testing

Combined testing generally refers to the integration of contractor and
government DT and government OT personnel on a single test team
(often referred to as a CTF). Previously, the contractor would
generally perform much of the DT, with dedicated government test
periods at the end of each major phase. Similarly, government opera-
tional testers had a largely independent test plan for OT. Now, with
integrated test teams or a CTF, government DT and OT personnel
are involved from the early planning stages through the completion of
all major test activities. Our contacts felt that the early involvement of
OT personnel in DT saved both costs and schedule for the program.
The statutory requirement for an independent organization to
do operational evaluation at the system level is now interpreted as
requiring only independent analysis of representative test data. Air
Force government-only testing (IOT&E) and Navy OPEVAL are
normally conducted in dedicated tests that the OT organization
performs, augmented, if necessary, by other service assets. With the
exception of any contractor logistics and maintenance support that
would be a normal part of the operation of the system, neither the
contractor nor the government development organization participates
directly.
In general, the advantages of combined test teams are
• elimination of redundant test activities
• early identification of issues and problems
• effective communication on potential efficiencies and work-

arounds throughout the test program.
Nearly all interviews indicated that program office, contractor,
and OT personnel see early involvement as positive and useful, mak-
ing such comments as “tremendous benefits in terms of schedule and
cost.” Ideally, the operational testers are directly integrated into the
CTF as full participants throughout the test program. The only nega-
tives mentioned were some potential for an adversarial relationship to
develop between the IPT representing the acquisition function
(which is therefore trying to field the system in the minimum time
and cost) and the operational testers representing end users and main-
tainers (who are attempting to maximize utility once the system has
been fielded). There was also some concern in the OT community
that informal opinions rendered early in the test program could be
perceived as definitive commitments by the OT activity. A consistent
concern on the part of both Air Force and Navy test communities
was the perceived shortage of qualified OT personnel. This often
limited their ability to support extensive early involvement, particu-
larly for smaller, less-visible programs. In some cases, AFOTEC has
hired support contractors to fill specialized slots for which Air Force
personnel are not available.
One of the contractor test managers observed that, because of
the degree of management integration in modern test programs, cost
analysts must be cautious about trying to isolate discrete elements of a
test program artificially, since the objective is to conduct many previ-
ously separate activities concurrently and thus gain synergy from the
integrated activities.
Multiservice testing involves systems that more than one service
will use. In these cases, which are becoming the norm for air-
launched weapons and, to a lesser extent, for aircraft, a lead service is
designated and has primary responsibility for executing the test pro-
gram. In general, requirements that are unique to one of the partici-
pating services are normally tested by that service as part of the overall
test program. The cost implications of multiservice testing depend on
the degree of commonality of requirements and configurations

among the participating services.
Contractor Versus Government Test Facilities

Another question we posed in our interviews was whether there had
been any consistent trend toward the use of either government or
contractor test facilities. This issue was raised because of economic
pressures on both the government and contractors—on the govern-
ment to close duplicative or low usage facilities and on the contrac-
tors to reduce fixed infrastructure costs. The dilemma the govern-
ment faces is the conflict between maintaining facilities that are
perceived as providing critical or unique test capabilities and
following a fundamental principle of acquisition reform, that the
program manager must have maximum flexibility to optimize his or
her test program. The first goal would lead to a policy mandating use
of government facilities for testing, while the second goal would leave
the program manager free to choose where to test. We were
specifically interested in whether the recent trend in some programs
of giving the contractor wide latitude in how and where the test
program will be executed resulted in significant shifts in test facilities.
Not surprisingly, the consensus from both the government and
contractor personnel was that the contractors would generally prefer
to do as much testing in their own facilities as possible. They feel that
this gives them greater control over cost and schedule by minimizing
travel costs, facilitating communication, and reducing charges for
infrastructure that does not add value to the product. Of course, few
contractors have open-air test ranges or certain specialized facilities
that they do not commonly use as part of their normal operations. As
a result, nearly all open-air testing for aircraft and guided weapons,
and for such specialized operations as climatic and electromagnetic
effects, electronic warfare (EW), carrier suitability, etc., is done at
government ranges and facilities. Contractor facilities are generally
used for M&S and system integration laboratories, since these are also
used for development activities. In general, contractors have the
capability to do most component and subsystem testing in their own

facilities. Hardware-in-the-loop testing can be done in either
government or contractor facilities, depending on circumstances and
availability.
While there was general agreement that the major government
test facilities are essential for executing the required test programs,
some contractor personnel expressed varying levels of frustration in
their dealings with the government test organizations. In programs
with aggressive affordability goals, there was a concern that some gov-
ernment test organizations were not as focused on controlling the
costs and schedule of the test program as other members of the test
team were. The respondents felt some practices at the test ranges were
overly conservative and caused unnecessary costs and delays. In some
cases, delays resulted from chronic understaffing or rigid procedures
with little provision for flexibility (particularly in areas perceived as
involving safety). A representative of one contractor noted that its
criteria for selecting among the available test facilities on a recent pro-
gram were, in order,
• cost
• responsiveness
• past performance
• capability.
Another contractor representative noted that the government
ranges tended to be “overfacilitized but undermodernized,” with
extensive (and often unique) infrastructures to support and limited
funds for investment in modernizing test equipment and facilities.
These issues are of increasing importance with TSPR contracts, in
which contractors are attempting to perform within what effectively
become fixed test budgets and schedules.
One of the challenges government test facilities face is that they
must recoup a high percentage of their operating costs through user
fees. All the ranges we visited spend considerable time and effort to
set these fees by projecting costs and usage for the coming fiscal year.
Although this is integral to financial management in industry, it is
more challenging in a government environment, in which managers
have limited control over some parts of their cost structure. In addi-
tion, although test ranges are, in theory, involved in test planning for
all major programs, there is a significant amount of schedule “churn.”
For example, at one range, half the actual testing for the year report-
edly was not included in the original schedule used for budgeting.
Despite these difficulties, both government and contractor program
personnel said significant unanticipated changes in range use rates
were fairly uncommon. When they did occur, the result was usually
reduction or deferral of the planned testing.
Given the occasional schedule turmoil, test article or range
equipment problems, weather delays, and limitations on range capac-
ity, surprisingly few programs complained of getting “bumped”
because of higher-priority users. This could have been due, in part, to
the fact that most of the programs in the study were relatively large or
high priority. We did hear of range availability delays of up to six
months for one FMS program.
Live-Fire Testing
Full-scale system-level live-fire testing (or a formal waiver certifying
that it would be unreasonably expensive or impractical and describing
proposed approaches for assessing vulnerability, survivability, and
lethality) has been a congressionally mandated requirement since
November 1986. Such testing has therefore been a part of all covered
programs since then. However, we found that the costs were often
combined with other DT activities.
Although the requirements for live-fire testing, especially for
modifications of previously fielded systems, are often hotly debated at
the service headquarters level, most program personnel seemed to
consider the testing to be a “fact of life” requirement and planned
accordingly. This could be because live-fire test plans and any waiver
requests must be submitted and approved early in the program
(before Milestone II or B), and are therefore integral to all subsequent
test planning. In the case of modification programs, the cost of live-
fire testing varies depending on whether the modifications are likely
to effect vulnerability, survivability, or lethality. Current programs

commonly make extensive use of M&S for vulnerability, survivabil-
ity, and lethality analysis, as well as for test design, to maximize the
cost-effectiveness of testing.
Warranties
None of the government or industry test personnel interviewed indi-
cated that warranties significantly changed the T&E process or costs.
The T&E scope was set independently of warranty considerations.
Cost Trends
In the preceding subsections, we examined trends in various aspects
of the T&E process. A key question remains: Have these initiatives,
in the aggregate, had the effect of reducing the overall cost of T&E?
Although it is difficult, if not impossible, to quantify the cost
effects of each of these in isolation, we did attempt to determine the
overall direction of T&E costs since the early 1970s. Although test
programs clearly differ in content, both individually and by system
type, it is at least apparent that costs are not trending downward. We
also examined the possibility that test durations might be shorter.
Figure 3.1 shows that, if anything, they are lengthening.
Have all the changes to the T&E process described above really
been ineffective? based on the interviews and some circumstantial
evidence, it appears that any net savings due to increased efficiencies
in testing are being offset by other trends.
Improvements in testing efficiency and effectiveness are difficult
to isolate and quantify from the available data. Nearly all the test per-
sonnel in government and industry commented on the increased pro-
ductivity made possible by advances in M&S data collection systems
and approaches and by the reduction of redundant testing between
contractor and government DT activities and between the DT and
OT test communities. While it is difficult to assess productivity
Figure 3.1
Flight-Test Duration of Aircraft Development Programs Over Time
Fighter/attack Bomber Cargo/trainer

100
90
80
Flight-test duration (months)
70
60
50
40
30
20
10
0
F-14 F-15 F-16 F/A- AV- F/A- F-22* B-1A B-1B B-2 C-5A C-17 V-22 V-22 T-45
18A/B 8B 18E/F FSD EMD*
NOTES: Measured from first flight to end of DT. By comparison, the Boeing 777 flight-
test program lasted only 11 months, during which nine test aircraft provided 69
aircraft months of testing and logged approximately 7,000 flight hours. Of course,
commercial aircraft do not require testing in the range of mission requirements and
flight regimes typical of military aircraft.
* Program in progress.
RAND MG109-3.1
improvements objectively, given the lack of consistent test activity

data over time, we did find comparisons in some areas. While these
should be considered to be examples rather than a definitive sample,
they tend to confirm much of the qualitative feedback on testing
trends we received from government and industry personnel.
The proliferation of digital systems on modern aircraft has
shifted much functionality (and complexity) from hardware to soft-
ware. In some ways, this has actually improved testability by shifting
test activities from mechanical, electrical, and hydraulic systems to
software and firmware, which not only can be tested more efficiently
in ground facilities but can themselves facilitate testing and data col-
lection. For example, aircraft fatigue testing previously involved
designing and building special test equipment to provide required

inputs (movements, deflections, and loads) to the test article. With
digital flight controls, precise inputs can be commanded through the
flight control system itself. Some aspects of digital flight control and
avionics performance can be monitored by recording data directly
from the digital data bus without the weight, space, and expense of
dedicated test sensors.
The amount of test data recorded is one indicator of the
increased scope of testing. Test equipment installed on the F/A-
18A/B would typically generate approximately 256 kilobytes of data
per flight. By comparison, the Joint Strike Fighter (JSF) is projected
to provide 3 to 4 gigabytes, much of it through the fighter’s own data
bus. A related improvement, one Boeing strongly emphasized, is that
test engineers participate in system design to address testability and
test instrumentation concerns. This allows for efficient installation of
provisions for instrumentation and wiring during design and test
article manufacturing rather than trying to do so later, on an ad hoc
basis.
However, despite these advances in the conducting of tests,
other changes appear to have offset any potential net savings. Most of
these changes can be assigned to one or more of the following catego-
ries:
• system complexity
• higher test standards
• increased test charges to programs.
Continuing advances in technology have translated into system
capabilities unimagined a generation ago. The growth in capability
translates, at least indirectly, into increased test complexity. Figure
3.2 shows the relative growth in the amount of flight-test data
collected for three representative fighter aircraft developed since the
early 1970s. Although it is normally simpler to collect digital data
than to collect the corresponding analog data, the figure makes it
clear that the amount of data to be analyzed has greatly increased.
Table 3.1 highlights the effects of the advanced capabilities of the
F-22 on testing.
Figure 3.2
Growth of Flight-Test Data Collected as Indicated by Number of Sensors or
Measurement Points
12,000
Digital
10,000 Analog
8,000
Measurands
6,000
4,000
2,000
0
F-15A/B F/A-18 A/B F/A-18 E/F
SOURCE: Boeing.
NOTE: Measurands are the number of sensors or measurement points.
RAND MG109-3.2
Table 3.1
How Advanced Capabilities Affect Testing
Feature Requires
Reduced-signature design Maintaining several test aircraft in low observability
(LO) configuration
Obtaining and scheduling unique LO test assets
Managing security
Internal weapon carriage Additional wind-tunnel characterization of the flow
field with the bay open
An additional flight-test configuration (doors open)
for performance and flying qualities
Sustained supersonic cruise Reducing test time with chase aircraft
Adding tanker support
Increasing the use of supersonic test airspace
Table 3.1—continued
Feature Requires
Thrust vectoring Special ground-test fixtures to control vectored
exhaust gases
Multiaxis force and moment instrumentation for
measuring thrust
Ground and in-flight performance testing
Expanded flying and handling quality testing
Failure modes and effects testing, particularly with
respect to asymmetric actuation
Integrated avionics Additional EMI and EMC testing
Comprehensive ground and air testing of integrated
system modes
Collecting data on system timelines and their effects
on system performance
Sensor fusion High-density, multispectral, integrated, enhanced-
fidelity target and threat simulation
Comprehensive integrated ground-test facilities
Highly integrated, wide- Multiple threat and target simulators with high
field-of-regard sensors update rates operating concurrently and having a
large field of view
Tailored countermeasures A target platform with a representative signature
Air and ground threats that appropriately stimulate
the system to determine countermeasure effec-
tiveness
Integrated maintenance First-of-kind DT/OT evaluations and assessments (for
information system and software-intensive, paperless systems)
technical order data
AFMSS/mission support A higher level of integration in the testing activity,
element because of the higher level of integration of these
systems in the weapon system
SOURCE: F-22 TEMP, 1999.
In addition to the obvious increases in the complexity of the sys-

tems under test, there is also some indication that the standards to
which current systems are tested are higher than those for legacy air-
craft and missiles. Table 3.2 compares the metrics from fatigue test-
ing of the F/A-18 A/B (developed in the late 1970s) with those for
the F/A-18 E/F (developed in the mid-1990s). (Despite the nearly
twelvefold increase in the data collected, the labor required dropped
by approximately half.)
Similarly, the F/A-18 A/B was certified for two weapon configu-
rations during DT; the F/A-18 E/F DT certified 29. Current plans
Table 3.2
Comparison of F/A-18 Fatigue Testing
Metric F/A-18A/B F/A-18E/F

a 615 1,643
Strain gauges at start
b 18 89
Deflections at start
c 132 1,560
Data channels
SOURCE: Boeing.
a
Represents the amount of instrumentation.
b
Represents the number of cycles.
c
Represents the data volume.
for the JSF are to certify more than twice as many weapons as the
F/A-18E/F. Current projections show that the JSF will have roughly
three times the 2.1 million lines of code projected for the F-22 air-
vehicle software.
Overlaid on the increasing complexity and scope of test pro-
grams is the increase in the fees for using DoD test facilities. As
Figure 3.3 shows, user fees pay a significant proportion of the cost of
operating test facilities. This is consistent with the fee-for-service shift
many DoD activities have made. While this does not change DoD’s
overall cost, spreading fixed costs among fewer users increases the
costs for individual programs, unless infrastructure costs can be
reduced proportionately.
Figure 3.3
Comparison of Institutional and User Funding for Major Air Force and Navy
Aircraft and Weapon Test Centers
250
Institution
User
200
Funding in FY 1999 ($M)
150
100
50
0
AF Flight NAWC Air Armament NAWC China
Test Center Patuxent River Center Lake/Pt. Mugu
SOURCE: Defense Science Board (1999).
RAND MG109-3.3
CHAPTER FOUR
Test and Evaluation Cost Data
This chapter discusses T&E cost data, addressing the following in

particular:
• T&E cost-element definitions
• how contractors and the government develop and report costs
• the cost data sources for this monograph
• cost data caveats
• how this analysis aggregated the cost data.
Cost Element Definitions

In any study of this type, it is important to define clearly the costs
that are being analyzed and those that are being excluded. Military
Handbook 881 (MIL-HNBK-881) provides guidance on developing
and using work breakdown structures (WBSs) in the development
and acquisition phases of DoD programs. Although the definitions
from the handbook we present below are largely verbatim, we have
condensed them and have omitted many of the examples in the
handbook. See Appendix C for the actual Section H.3.3 (ST&E)
from MIL-HNBK-881.
System Test and Evaluation

ST&E is the use of prototype, production, or specifically fabricated
hardware or software to obtain or validate engineering data on the
performance of the system during the development phase of the pro-
59
gram. ST&E includes detailed planning, conduct, support, data

reduction and preparation of reports from such testing, and all hard-
ware and software items that are consumed or planned to be con-
sumed during such testing. ST&E also includes all effort associated
with the design and production of models, specimens, fixtures, and
instrumentation in support of the system-level test program.
This category does not include test articles that are complete
units; these are funded in the appropriate hardware element. So, for
example, the cost of manufacturing flight-test air vehicles should be
included in the Air Vehicle WBS element, while static, fatigue, and
drop test articles are included in ST&E.
The handbook divides ST&E into five main elements:
• development T&E (DT&E)
• OT&E
• mock-ups
• T&E support
• test facilities.
Development Test and Evaluation

DT&E is planned and conducted and/or monitored by the DoD
agency developing the system. The ultimate purpose of DT&E is to
demonstrate that the development process is complete and that the
system meets specifications. The outputs are used throughout devel-
opment to support performance evaluation, trade-off studies, risk
analysis, and assessments of potential operational utility. DT&E
includes wind tunnel, static, drop, and fatigue tests; integration
ground tests; test bed aircraft and associated support; qualification
T&E, developmental flight tests, test instrumentation, and avionics
testing. Table 4.1 shows representative contractor activities normally
included in DT&E for aircraft and guided weapons.
Operational Test and Evaluation

OT&E is the T&E that agencies other than the developing command
conduct to assess the prospective system’s military utility, operational
Test and Evaluation Cost Data 61
Table 4.1
Representative Contractor DT&E Activities for Aircraft and Guided Weapons
Aircraft Guided Weapons

System test requirements and planning System test requirements and planning
M&S M&S
Wind-tunnel tests Wind-tunnel tests
Static article and test Structural tests
Fatigue article and test Environmental tests
Drop article and test Special test articles
Subsystem ground tests Other ground tests
Avionics integration tests Flight-test support
rmament and weapon delivery
integration tests Test aircraft preparation
Contractor flight test Telemetry kits
Special testing Targets
Other T&E Other T&E
T&E support T&E support
effectiveness, operational suitability, and logistics supportability.

OT&E includes any contractor support used during this phase of
testing.
Mock-Ups
Mock-ups encompass the design engineering and production of sys-
tem or subsystem mock ups that have special contractual or engi-
neering significance or that are not required solely for conducting
either DT&E or OT&E.
The reported costs for mock-ups have been only a small part of
total ST&E for recent programs, and the relative costs of mock-ups
and the share of total ST&E costs appears to be declining over time.
This decline may be due to advances in computer modeling that
reduce the need for elaborate physical mock-ups. It may also be that
most mock-up costs tend to be reported under DT&E rather than
separately.
Test and Evaluation Support

T&E support includes the effort necessary to operate and maintain,
during T&E, systems and subsystems that are not consumed during
the testing phase and are not allocated to a specific phase of testing. It
includes, for example, spares, repair parts, repair of reparables, ware-
housing and distribution of spares and repair parts, test and support
equipment, test bed vehicles, drones, surveillance aircraft, contractor
technical support, etc. We found that the content of T&E support
varies considerably across programs.
Test Facilities
Test facilities here are the special facilities required to perform the DT
necessary to prove the design and reliability of the system or sub-
system. These facilities include white rooms and test chambers but
exclude brick-and-mortar facilities identified as industrial.
In general, program offices fund only the test facilities that are
unique to their program. Once a facility is available, succeeding pro-
grams may arrange to use the facility and thus pay some of the vari-
able costs. It can be difficult to predict the program office’s share of
such costs because the required test facilities may be available as part
of the existing infrastructure; may be funded by direct test infrastruc-
ture funding; or may require funding from the program office, either
alone or in conjunction with other users. It is noteworthy that the
B-2 and F-22—programs with unprecedented levels of sophisticated
avionics, LO technology, and high levels of security—dedicated con-
siderable portions of their T&E budgets to paying for special facilities
that were not available as part of the existing test infrastructure.
Cost Collection and Reporting

The government program manager is responsible for developing the
new system and decides how to acquire each of its elements. At the
outset of a program, the program manager defines a WBS that repre-
sents the system and supporting activities in a product-oriented hier-
archy consisting of hardware, software, facilities, data, services, and
other work tasks. This hierarchical structure completely defines the
system and the work to be done to develop and produce it. MIL-
HDBK-881 contains generic three-level WBSs, organized by com-
modity type (including aircraft and missiles), to provide program

managers a starting point for defining the WBS for a specific pro-
gram. The program manager usually places certain WBS elements on
contract and requires the contractor to report costs according to the
defined WBS.
Contractor Costs
Contractors report their costs to the government using a variety of
standard reports, primarily the cost performance report (CPR) and
variations of the contractor cost data report (CCDR). The source of
the cost data for these reports is the contractor’s accounting system.
Since the cost account structure of most contractors’ accounting sys-
tems will not match the approved program WBS, the costs are allo-
cated or grouped into the government-approved program WBS for
reporting purposes (providing a crosswalk between the accounting sys-
tem and the cost report).
The DoD has used this well-defined and accepted WBS struc-
ture, and standardized cost reports tied to it, for decades. This has
resulted in a large body of historical contractor costs that are intended
to be comparable within a given commodity type. The degree of
comparability depends on how well the crosswalk has been done,
which itself depends on the degree of difference between the contrac-
tor’s account structure and the program WBS and on the amount of
effort that has been devoted to reconciling them. Because of the long-
standing requirement for cost reporting and the routine use of the
data, for both management of current programs and analysis of future
programs, the system generally produces a useful high-level picture of
program costs. The fidelity and/or consistency of cost reporting tends
to decrease at lower levels of the WBS, for smaller programs, with the
extensive use of integrated product (or process) teams, and when
commercial practices are used as a justification for reducing or elimi-
nating contractor cost reporting.
Government Costs
In contrast to the long-standing and consistent method for capturing
and reporting contractor costs associated with weapon systems, the
methods used to capture and report government costs associated with

weapon systems vary across organizations. Government financial
reporting systems vary in their accuracy, categorization, comprehen-
siveness, level of detail, and availability. We found no central
repository of government test costs at the military service or systems
command levels. Rather, costs were captured, reported, and stored at
the organizations directly responsible for testing and/or in the indi-
vidual program offices. The following paragraphs describe the fund-
ing sources and rate structures for test activities and how government
costs are captured and reported at the organizations responsible for
testing, based on our visits to several such organizations.
The rules governing the funding of test organizations, the proc-
esses by which they are funded, and their rate structures are complex,
and the details are beyond the scope of this monograph. These para-
graphs provide a context from the perspective of users of test facilities.
Test organizations receive funding from a number of sources. For the
purposes of this report, the funding sources fall into two categories:
users of the facilities and all other sources. The primary users of the
facilities include DoD program managers; contractors working on
DoD contracts; foreign military sales (FMS) organizations; and, to a
lesser extent, non-DoD contractors, state and local governments, and
private parties. In addition to users, the other sources of funding
include appropriated funds classed as institutional or “direct budget
authority.” These funds are intended for maintaining and upgrading
the general-purpose test infrastructure.
All the test facilities we visited and from which we collected
costs are part of the Major Range and Test Facility Base (MRTFB).
The MRTFB comprises major test activities regarded as national
assets that receive institutional funding from DoD. The test activities
are required to use a uniform reimbursable funding policy, under
which users are charged for direct costs, such as labor, materials,
equipment, and supplies, and pay a proportional share of equipment
maintenance costs related to their use. In practice, the direct costs and
proportional share of costs charged to users are affected by the
requirement for the test activity to break even. The test activity must
balance its costs against institutional funding and customer funding
and must adjust the rates it charges customers so that the costs they
incur and the funding balance.1 The test facilities expend consider-
able effort avoiding unplanned rate increases because of the obvious
problems for users’ test budgets. Although there are differences in the
way Navy and Air Force test facilities are funded, the funding policies
and practices at the MRTFB facilities we visited were similar because
they are all governed by the same DoD financial regulations. It
should be noted that these funding practices apply to DoD facilities
and not to the National Aeronautics and Space Administration or
private facilities that DoD occasionally uses for testing.
In summary, the user organization pays for the costs of labor
and materials related to the tests conducted at the facility and pays a
proportional share of other test-related costs. The user’s rates are
affected by how much other business is at the facility that year.
The funding practices for OT vary more significantly between
the Navy and Air Force. For the Air Force, AFOTEC has its own
program element and pays for significant items, including range costs,
rental of equipment, contractor support, special supplies and equip-
ment for data reduction, civilian pay, per diem and travel, supplies
and equipment, and pretest planning. For the Navy, OPTEVFOR
pays for its own personnel and travel; the system program manager
pays for everything else. After OPTEVFOR develops a program OT
budget, the Navy program manager sends funds to OPTEVFOR for
its contracting office to arrange for range time, assets, additional con-
tractor support, etc. This difference in funding, although a relatively
small part of the total T&E costs for a weapon system, would tend to
make OT costs for a Navy program somewhat higher than for a
similar Air Force program, all else being equal.
The process of estimating government test costs for a program
begins when the program office identifies the need for testing. In the
Air Force, the SPO prepares a program introduction document (PID)
that identifies the system that needs to be tested, the test services and
_____________
1 We attempted to determine whether there had been significant shifts of costs to users to compensate
for shortfalls in institutional funding at the test centers but, because of the lack of consistent data over
time, could draw no conclusions.
test equipment required, the starting and ending dates, and other
related information. The program office sends the PID to the organi-
zation that will conduct the testing. The test organization responds to
the PID with a statement of capability (SOC). The SOC confirms
that the test organization has the resources to perform the testing and
contains a summary of proposed test events and resources, a schedule,
and costs. The SOC also provides an assessment of technical, sched-
ule, cost, and programmatic risks. The program office consents to the
terms in the SOC by providing funding and written confirmation to
the test organization.
The test organizations we visited track budgeted funding, obli-
gations, and expenditures, generally by organization or function.
However, in contrast with the tracking of contractor costs, there was
no consistent product or output or related WBS for government costs
across programs, so comparison across programs below the level of
total government costs is not currently practical.
The Navy’s process is not as formal as that of the Air Force.
Financial obligations and expenditures are typically tracked and
recorded in the program offices that originate the funds and in the
test activities. However, the Naval Air Systems Command (NAVAIR)
cost analysis group’s attempts to extract T&E cost data for this study
from the Navy’s Standard Accounting and Reporting System
(STARS) were generally unsuccessful.
As in the Air Force, the lack of a standardized WBS for govern-
ment costs results in a wide disparity in the level of detail of cost data.
For example, in some Navy programs, STARS could identify only
total funds by appropriation by fiscal year. Below the appropriation
level, identifying the amount spent on system T&E was impossible.
A more significant difficulty is that many government organiza-
tions dispose of or archive the cost data, so that it is often unavailable
a few years after expenditure. During budget execution, most pro-
grams can track their expenditures by activity and specific tasks.
However, after the funds have been expended, this information does
not appear to be systematically retained for future analysis. In some
cases it is archived but, it is, for all practical purposes, unavailable. In
other cases, it is simply disposed of. The notable exception was the
Earned Value Cost Analysis System in the 46th Test Wing at Eglin
Air Force Base, which is used for both management and analysis.
Costs for OT were, however, available for both services from their
OT agencies.
Cost Data Sources

We collected costs for use in this report from a variety of sources. For
contractor costs, we generally collected and used CCDRs. These
reports have the advantage of providing all the contractor’s costs on a
contract in an approved, uniform format. We collected government
costs either from the cognizant program office or from the test orga-
nizations involved, using whatever records were available to them.
Using the program office as the source of cost data has the advantage
that these costs should include all expenditures of program funding,
regardless of executing activity. The disadvantage is that these costs
are not consistently collected or retained. Information from test
activities is generally more detailed but may exclude funding and
effort on the program at another test activity.
Table 4.2 lists the sources from which we obtained contractor
and government costs for the aircraft programs. Similarly, Table 4.3
lists our sources for contractor and government costs for the guided-
Table 4.2
Aircraft Cost Data Sources
Contractor Costs Government Costs
Program Accounting Program Test

Office System Office Activity
CCDR/CPR Records (STARS) Records Records
B-1 CMUP X X X
B-2 X X X X
C-17 X X X
F/A-18E/F X X X
F-22 X X X
T-45 X X X
V-22 X X X
Table 4.3
Guided-Weapon Cost Data Sources
Contractor Costs Government Costs
Program Program Test

Office Office Activity
CCDR/CPR Records Records Records
AMRAAM FSD X X X X
AMRAAM Ph. 1 X X X
AMRAAM Ph. 2 X X X
AMRAAM Ph. 3 X X X
AIM-9X X X X X
JASSM X X X
JDAM X X
JSOW X X X X
SFW X X X
SLAM-ER X X
WCMD X X
weapon programs. In addition to these programs, our CER develop-

ment data set included data from selected legacy programs.
Cost Data Caveats

We made every effort to use credible sources of cost information and
to check the information for accuracy and completeness. However,
because of the different sources of costs and their varying quality and
reliability, we have less confidence in some data points than in others.
In addition, estimators who are using this information need to be
aware that several of the programs have peculiarities that affect their
costs. We therefore offer the following caveats:
• The B-1A program was cancelled before production, but flight
testing continued at a low level for several years before the B-1B
program began. Flight-test costs for the B-1A are higher than
those of other programs for the number of flights, flight hours,
and flight months and may represent additional effort in the
interim period between the formal development programs.
• The B-1B airframe was derived from that of the B-1A, so the
B-1B test program had less ground testing than would a normal
development program. For CER development, we combined the
costs for both test programs.
• The F-22 program was in progress as of this writing and is
included for information only. It was not used in any quan-
titative analysis because the content and cost of the remainder of
the flight-test program were uncertain. The costs shown were
developed from the estimate at completion (EAC) for flight test
from the November 2001 CPR that the program office
provided, plus the EAC for all other test elements from the
March 2001 CCDR.
• The proportion of ST&E costs relative to the total development
program is much smaller for the F-22 program than it is for any
other aircraft in the data set. There are at least two probable
reasons. One is that the program was in progress as of this
writing, and the final test cost may be considerably higher than
the current estimate at completion. The other possibility is
related to the F-22 program’s IPT organization and time-
charging practices. Contractor personnel who are primarily
assigned to a nontest IPT but who work on test tasks charge
their time to their original (nontest) IPT. This understates the
effort expended on testing relative to other programs.
• As Chapter Three discusses, the T-45 test program was executed
primarily by the contractor, with short periods of dedicated
government testing. This minimal level of government
involvement contrasts with the other programs for which we
have government costs, in which the government had a larger,
continuous presence and a more-active role. The government
test costs for the T-45 program are a much lower proportion of
total DT costs than for other programs in the data set.
Furthermore, the source of these costs is the Navy’s STARS,
which may not fully reflect all government test costs.
• The V-22 EMD test program was in progress as of this writing
and is included for information only. The content and cost of
the remainder of the program are uncertain. We used the EAC
from the December 2001 CCDR. By that date, the originally

planned test program was virtually complete. The current test
program is event-driven and could continue for an additional 30
months.
• The AIM-9X DT program was largely complete as of this
writing. We used an estimate at completion through fiscal year
2002 provided by the program office.
• Advanced Medium-Range Air-to-Air Missile (AMRAAM) Phase
3 is being executed on a TSPR contract, which gives the con-
tractor more latitude and responsibility in executing the con-
tract. This contracting arrangement results in higher contractor
test costs relative to the total cost because the contractor
arranges and pays the government entities for testing and costs
that the government would traditionally pay directly.
• The Joint Standoff Weapon (JSOW) Baseline (AGM-154A) was
used for analysis. An additional JSOW variant (AGM-154B)
was added to the original JSOW development contract as a con-
current modification; it has not gone into production. Program
cost reporting combined both variants, which made segregating
Baseline costs problematic. Although NAVAIR and the Air
Armament Center attempted to identify Baseline costs for us,
the cost data should be used with caution.
Many programs, especially missile programs, have test-related
efforts that are not reported as part of system T&E. The reporting of
test-related costs can vary from one contract to another, even for
modifications to the same basic missile by the same contractor. For
example, simulation costs were reported in ST&E in one AMRAAM
modification program, but in different elements in the other
AMRAAM modification programs. In general, the distinction
between component-level testing and system-level testing can be
unclear, and contractors use their best judgment in allocating costs.
Although we are aware that there are inconsistencies among reported
costs, we have not tried to adjust the costs as reported because we lack
consistent insight into all programs. The AMRAAM contracts are
unusually detailed and allow greater insight than most. Rather than
adjust the few programs into which we had sufficient insight and
leave the others alone, we chose to leave the costs for all programs as
reported.
How This Analysis Aggregated Cost Data

We had several objectives in working with the cost data. Our primary
objectives were to
• collect cost data representative of current T&E practices on
recent programs
• identify cost trends over time
• identify drivers of system test costs
• develop cost-estimating methodologies useful for cost estimators
trying to project resources at the beginning of a development
program.
Achieving these objectives was difficult because the two data sets
(for aircraft and guided weapons) were diverse, as was the quality of
the data. The aircraft data set includes aircraft of various types (fight-
ers, bombers, transport, and trainers) that require different kinds of
testing and have different development and unit costs. To allow trend
analysis, we included data on a variety of older aircraft.
Similarly, the guided-weapon data set includes weapons of vari-
ous types and various degrees of new development. The data set
includes unpowered weapon kits, such as the Joint Direct Attack
Munition (JDAM) and the Wind-Corrected Munitions Dispenser
(WCMD), as well as air-to-air and air-to-ground missiles. The scope
of development efforts ranges from missile modification programs to
full development programs. As with aircraft, weapons of different
types and with different amounts of development require different
kinds and amounts of testing.
The quality of the data also affected the way we aggregated them
and our ability to pursue questions analytically. Contractor costs for
aircraft ST&E had the highest level of detail in general, but the WBS
elements were not always consistent. Contractor costs for guided-
weapon ST&E ranged from great detail for the AMRAAM programs
to a single cost number for other programs. Government costs varied
similarly.
In light of these realities of the nature of the cost data, we
• identified cost trends over time through discussions with experi-
enced government and contractor test personnel and examined
cost trends at the total ST&E level by including legacy programs
• identified drivers of system test cost at a high level, consistent
with the quality of the data and diversity of programs, and
aggregated data accordingly
• aggregated data so that costs could be explained by variables
typically available to estimators at the beginning of a develop-
ment program.
We grouped aircraft contractor T&E costs into ground test,
flight test, other test costs, and a subset of ground-test costs consisting
of static and fatigue test costs. These groupings allowed identification
of differences in the cost of each group by aircraft type and by charac-
teristics of the test program. When they were available, government
T&E costs for aircraft had to be combined into a single figure
because they were generally provided to us in an aggregation that did
not allow consistent identification of tasks.
Contractor and government costs for guided weapons were each
combined into a single figure for analytic purposes. This was done for
two reasons. First, in contrast to the case with aircraft, the majority of
T&E costs for most guided-weapon programs are for government
activities. However, the nature of the T&E effort done by the
government varies from program to program, so only a total of both
costs consistently represents the entire test effort. Second, aggregation
at this level is more consistent with the objective of developing CERs
for analysts with limited knowledge of the program at the beginning
of a development effort.
After aggregating the cost data as described above, we attempted
to generate CERs for each aggregation. CERs were developed for air-
craft contractor ST&E costs both in total, as well as ground, flight,
and other tests. A separate CER was also developed for static and
fatigue testing. Unfortunately, we were unable to obtain reliable gov-

ernment T&E cost data for the legacy aircraft programs that were
used to supplement the data collected as part of this study. As a
result, there were too few observations of government costs over the
various aircraft types to develop CERs for government costs. (The
available government cost data appear in the limited-access technical
report.)
Because of the diversity of the weapon development programs,
we were unable to generate satisfactory CERs for the entire group of
weapon programs. However, we were able to develop satisfactory
CERs for the guided missiles. Our CER development efforts and
results are discussed in Chapter Five.
CHAPTER FIVE
Estimating Test and Evaluation Costs
The estimating resources in this report are intended to address two

situations. The first is when an estimate is needed early in the process
of defining a test program, before detailed planning information is
available. For example, estimators trying to project the amount of
resources for programming or budgeting purposes before SDD may
find the methodologies useful. The second is as a cross check to assess
the adequacy of resources planned, programmed, or budgeted for a
test program by comparison with actual costs on similar programs.
This chapter discusses uses of the data we collected for cost
analysis. The following topics are addressed:
• approaches to estimating T&E costs
• data analysis
• CERs.
Approaches to Estimating Test and Evaluation Costs

Because T&E is a complex set of activities tailored to each program,
no single estimating method will give the definitive answer. Cost esti-
mators generally use one or more of the following estimating
approaches:
• analogy to a similar program
• parametric CERs derived from multiple historical programs
• summations of costs estimated at a detailed level.
75
The data and methodologies in this monograph support the first two
approaches.
For estimating by analogy, the test program descriptions and the
limited-access supplement can be used to develop either a cross-check
or a primary estimate. If several programs are found to be similar to
the one being estimated, the analyst can use the data to develop cus-
tom CERs for the case being estimated. For those without access to
the supplement, Table 5.1 gives average values for the costs of com-
mon T&E elements and other characteristics of potential interest for
four representative multiengine fighter programs. Unfortunately there
were not enough consistently categorized data to develop similar
tables for other types of aircraft or for guided weapons.
For estimating by CERs or factors, we derived quantitative rela-
tionships when sufficient data were available and satisfactory relation-
ships could be developed. The aircraft and missile databases contain
diverse sets of programs of varying types, cost, and complexity. Obvi-
ously, the CERs can be most useful when the estimated program is
similar to those in the database.
Estimating by detailed buildup is normally done by test person-
nel who have knowledge of the test activities to be estimated and the
resources that should be required. These estimates are generally
developed by the test activity, program office, and system contractor.
In general, we recommend that estimators outside of test organi-
zations use a combination of the analogy and CER approaches to
estimate a test program. Because each test program is inherently
unique, CERs cannot reflect all the peculiarities of a given test pro-
gram but can provide an appropriate benchmark for similar pro-
grams. Carefully chosen analogies may be the best approach when
similar historical programs can be found. Even when there are signifi-
cant differences, useful comparisons to the historical data can often be
made. At a minimum, such comparisons can provide either a cross-
check or a starting point for a more in-depth assessment.
Other metrics, such as flight-test hours (Figure 5.1), average
flight-test hours per test aircraft month (Figure 5.2), and number of
guided-weapon launches per month (Figure 5.3), can also be helpful
Estimating Test and Evaluation Costs 77
Table 5.1
Average Test Costs for Representative Multiengine Fighters
(contractor system T&E FY 2001 $M)
Coefficient %
Standard of of
a
Average Deviation Variation ST&E
Nonrecurring development 2,595.8 494.7 0.19
Nonrecurring development
(less ST&E) 1,815.1 451.2 0.25
System T&E 795.9 59.0 0.07 100
Wind tunnel test program 37.8 14.3 0.38 5
Static articles and tests 57.6 31.3 0.54 7
Fatigue articles and tests 42.6 6.5 0.15 5
Drop and accelerated loads tests 24.7 1.5 0.06 3
Air vehicle subsystem test program 55.8 14.7 0.26 7
Avionics integration tests 43.4 13.5 0.31 5
Armament and weapon integration
tests 7.4 4.0 0.54 1
Mockups 20.3 11.5 0.57 3
Other ground 11.9 8.4 0.71 1
Ground test subtotal a 267.4 20.5 0.08 34
Contractor flight tests 367.8 33.4 0.09 46
Other test subtotal 160.8 54.2 0.34 20
Weight empty (lbs.) 28,998 6,203 0.21

Aircraft months 263 28 0.11
Test duration (months) 46 9 0.19
Flight hours 3,771 1,164 0.31
Flight hours per aircraft per month 14.3 3.9 0.28
Contractor ST&E spent by first flight
(%) 44.8 11.2 0.25
a
Note that the ground-test subelements do not sum to the ground-test subtotal. In the
few cases for which subelements for a particular aircraft had zero values, they were
excluded from the calculation of the averages to avoid distorting the averages of the
individual subelements.
for attempting to assess the realism of a planned test program based

on the experience of previous programs.
In developing parametric CERs, we began with a visual analysis
of plots and graphs, then used statistical analysis (linear regression) to
develop the CERs. We had three criteria for selecting explanatory
variables to predict ST&E costs:
Figure 5.1
Total DT Flight Hours
Fighter/attack Bomber Cargo/trainer
7,000
6,000
5,000
Flight time (hours)
4,000
3,000
2,000
1,000
0
F-14 F-15 F-16 F/A- AV- F/A- F-22* B-1A B-1B B-2 C-5A C-17 V-22 V-22 T-45
18A/B 8B 18E/F FSD EMD*
NOTE: * Program in progress.
RAND MG109-5.1
Figure 5.2
DT Flight Hours per Aircraft Month
Fighter/attack Bombers Cargo

35
30
25
Flight time (hours)
20
15
10
0
F-14 F-15 F-16 F/A- AV- F/A- F-22* B-1A B-1B B-2 C-5A C-17
18A/B 8B 18E/F
RAND MG109-5.2
Figure 5.3
Guided DT Launches per Month in Order of Contract Award Date
Air-to-air Air-to-ground/cruise
1.4
(94)
1.2
Launches per month (number)
1.0
0.8 (13)
(27) (27)
0.6
(20)
0.4 (10)
(8)
0.2 (12)
(6)
0.0
AMR- AMR- AMR- AIM- AMR- HARM JDAM SLAM- JAS-
AAM AAM AAM 9X AAM 1978 1995 ER 1995 SM*
1981 Ph 1 Ph 2 1996 Ph 3 1998
1991 1994 1998
Number of DT guided launches
RAND MG109-5.3
• The explanatory variable must have a logical relationship to cost

based on the research we had done for the study.
• It must, in general, be available to or determinable by estimators
early in a development program.
• It had to be statistically significant in the regression equation.
In addition to these criteria, we attempted to select variables that
could be useful for estimating modification programs and for esti-
mating new development. 1
The data we used for CER development included some legacy
programs in addition to the more-recent programs collected for this
_____________
1 Unfortunately, we had only limited success in obtaining usable T&E cost data on aircraft modifica-
tion programs (the B-1 Conventional Mission Upgrade Program [CMUP] and the T-45 Cockpit-21
program were exceptions), so we derived the CERs from full development programs.
study. Limiting the data set to recent programs would not have pro-
vided sufficient data for meaningful CERs. Expanding the data set to
include older programs seemed reasonable because the relationships
between our selected cost drivers and the contractor T&E costs of
legacy aircraft programs appear to be generally consistent with more
recent data. Unfortunately, while government costs were collected for
the programs included in this study, they were generally not available
for the legacy aircraft programs. As a result, the aircraft CERs include
contractor costs only. The cost behavior of weapon programs also
appeared roughly consistent over time, so our data set included
selected legacy programs.
A Priori Expectations of Aircraft ST&E Relationships

These criteria resulted in a short list of candidate variables for aircraft
ST&E costs. Table 5.2 summarizes aircraft test program characteris-
tics. For aircraft ground-test costs, we expected the weight and/or
unit cost to be good candidates. We expected weight to be the best
predictor of static and fatigue test costs.
These tests require building ground-test articles that are struc-
turally representative of the aircraft, which is itself costly. A cagelike
structure is built around the test article to hold the controls and
instruments required to conduct the structural tests. Fatigue testing
applies representative forces to the airframe structure in repetitive
cycles to simulate accelerated structural aging. These tests may last for
over a year. The test equipment for a large cargo aircraft is the size of
an aircraft hangar and weighs tons. Weight is commonly used as a
good predictor of the cost of manufacturing an airframe and is clearly
related to the test equipment used in this subset of tests.
Other ground-test costs include subsystem, avionics, and pro-
pulsion integration tests. We expected the cost of the airframe and
these subsystems to drive test costs. Because the cost of the airframe,
Table 5.2
Summary of Aircraft Test Program Characteristics
Average
Flight
Flight Hours per First
Aircraft Hours Aircraft Empty Flight
a
Months Duration (DT) Month Weight Year
F-14A 250 45 2,685 10.7 36,825 1970
F-15A/B 253 59 2,856 11.3 26,250 1972
F-16A/B 156 25 2,581 16.5 13,371 1976
F/A-18A/B 305 40 4,922 16.2 22,351 1978
AV-8B 147 37 2,038 13.9 12,931 1981
F/A-18E/F 244 41 4,620 18.9 30,564 1995
F-22 N/A N/A N/A N/A 31,670 1997
B-1A+B 264 131 3,425 13.0 181,400 1974/
1984
B-2 310 95 5,032 16.2 153,700 1989
C-5A 108 61 3,145 29.2 320,085 1968
C-17 169 39 4,104 24.3 269,696 1991
V-22 FSD N/A 43 763.6 N/A 31,886 1989
V-22 EMD N/A N/A N/A N/A 33,140 1997
T-45 N/A 64 1,932 N/A 9,394 1988
B-1 CMUP 32 11 712 22.3 181,400 1997
a
Months from first flight to end of DT.
subsystems, and avionics may not be available separately at the begin-

ning of a program and because these aircraft elements constitute
much of the cost of an aircraft, the unit cost for the air vehicle
(excluding propulsion cost) is a reasonable substitute. 2
We expected aircraft flight-test costs to have both fixed and vari-
able components. A flight-test program requires a certain amount of
staff and equipment infrastructure. Once the infrastructure is in
place, its costs are incurred regardless of whether or not the aircraft
are flown. These costs are relatively fixed and cannot be readily
turned on and off in response to short-term variations in test activity.
In addition to the fixed costs, there are variable costs for personnel,
test and support aircraft, and such expendables as fuel and parts.
Thus we expected that the best explanation of the costs of flight test-
_____________
2 We calculated the theoretical first unit (T ) cost using an 80-percent unit theory cost-improvement
1
curve. Note that only data from FSD/EMD lots were used to develop T1 costs.
ing would be some measure of the fixed nature of the flight-test

effort, such as length of DT, and some measure of the amount of
flying, such as number of flights or flight hours, or a variable that
captures both fixed and variable aspects, such as test aircraft months.
It was more difficult to find an intuitive cost driver for aircraft
“other test” costs. One reason for the difficulty is that this group of
costs includes rather amorphous WBS elements found in most air-
craft development programs, such as T&E support and other system
T&E. These are various T&E costs that cannot be clearly assigned to
either ground or flight-test activities. In attempting to explain this
miscellaneous group of costs that are related to both ground and
flight-test events, we expected them to be a function of both ground
and flight-test activity or of nonrecurring development cost.
In recognition of the argument made by some test managers that
test programs should be a set of integrated and synergistic efforts that
should not be arbitrarily separated or considered to be isolated, inde-
pendent activities, we also developed CERs for ST&E at the total
contractor cost level. We again selected explanatory variables, such as
weight or unit cost, that seem to drive ground tests and variables,
such as flight hours, aircraft months, or DT duration, that we expect
drive the fixed and variable flight-test costs. We also examined nonre-
curring development cost as a measure of the overall complexity of
the development effort and, by implication, of the test program.
Results of Aircraft ST&E Analysis

Using these variables that met our first two criteria of having a logical
relationship to test costs and being generally available to cost estima-
tors, we obtained reasonable statistical relationships for total ST&E,
ground test, static and fatigue test, flight test, and other test costs.3
Definitions and abbreviations of the variables are shown in Table 5.3.
_____________
3 The values of the input parameters for many of the CERs may change as the development program
proceeds. When estimated parameter values are used, they should be varied over an appropriate range to
examine sensitivity in the predicted values.
The correlation matrix for the variables is shown at Table 5.4 and
uses the same abbreviations.
The CERs presented below as “preferred” best met our selection
criteria. In addition, we developed alternative CERs using different
independent variables, formulations, or a subset of the available data
for use in estimating situations when these constructions may be
more appropriate.4 All the CERs have F-statistics for the equation
and t-statistics for the independent variable that are significant at the
95-percent confidence level,5 unless otherwise noted. They express
contractor costs in constant FY 2001 million dollars.
Table 5.3
Definitions of Aircraft Variables
Variable Variable Definition Units

NR DEV (Nonrecurring development less test) Contractor’s FY 2001 $M
total cost for the nonrecurring portion of the
aircraft development program, excluding the cost
of ST&E
STE (Contractor ST&E) Contractor’s total cost for ST&E, FY 2001 $M
excluding the government’s cost for DT or OT
GNDa (Contractor ground test) Costs for contractor FY 2001 $M
ground testing (wind tunnel, static, fatigue, drop,
subsystem, propulsion integration, avionics
integration, weapon integration, mockups, and
similar tests)
FLTa (Contractor flight test) Costs for contractor flight FY 2001 $M
testing (includes only events explicitly labeled as
flight tests in the contractor’s cost reports)
GND + FLTa (Contractor ground and flight test) Contractor’s FY 2001 $M
total cost for ground and flight testing, as defined
above
OTHERa (Contractor other test) Contractor’s T&E cost for FY 2001 $M
tests not included in ground or flight testing (e.g.,
T&E support and test requirements)
_____________
4 As previously noted, the lack of government T&E cost data for legacy programs precluded develop-
ment of CERs for government costs.
5 F- and t-statistics are measures of the significance of the coefficients of the entire equation and the
individual coefficients, respectively. Adjusted r2 is a measure of the variance “explained” by the selected
equation, adjusted for the degrees of freedom. The coefficient of variation is the estimated standard error
of the regression equation divided by the mean of the dependent variable.
Table 5.3—continued
Variable Variable Definition Units
ST/F a (Static and fatigue) Contractor’s T&E costs for static FY 2001 $M
and fatigue testing; includes the cost of building
the test articles and conducting the tests
T1 (T1 cost) Theoretical first unit (T1) cost for the flight- FY 2001 $M
test units built during the development programb
WEc (Weight empty) Total weight of the aircraft Pounds
structure and its subsystems, avionics, and engine
AC Mo (Aircraft months) Total time each flight-test aircraft Months
is available for flight testing during DT
EMD DUR (EMD duration) Elapsed time from development Months
contract award to end of DT
DUR (Flight test duration) Elapsed time from first flight Months
to end of DT
FLT HRS (Flight hours) Accumulated flying time during DT Hours
for all flight-test aircraft
F/A (Fighter/attack) Dummy variable used in regression 1 = fighter/
analysis to distinguish among different classes, in attack
this case, between fighter or attack aircraft and aircraft
other aircraft 0=
non–fighter/
attack
aircraft
CGO (Cargo) Dummy variable used in regression analysis 1 = cargo
to distinguish among different classes, in this case, aircraft
between cargo and other aircraft 0=
non–cargo
aircraft
a
Subtotals.
b
Calculated by dividing the total recurring air vehicle cost in development by the
number of FSD and EMD units, assuming an 80-percent cost improvement curve, using
the curve to calculate an algebraic lot midpoint, and backing up the curve from that
point to the T 1 cost:
Average unit cost

T1 cost =
(Lot midpoint) -0.322
Lot midpoint formulae can be found in various cost estimating or economics texts.
c
According to MIL-STD-1374 (DoD, 1977), weight empty is more inclusive than either
structure or airframe unit weight.
The data set for the CERs contains the following aircraft devel-
opment programs: AV-8B, B-1A and B-1B (combined), B-2, C-5A,
C-17, F-14, F-15, F-16, F/A-18A/B, F/A-18E/F, and T-45. The
AV-8B and F-16 programs include cost and programmatic informa-

tion from their prototype programs. We excluded the F-22 and V-22
programs from the data set used for generating CERs, even though
we collected information on the programs, because they are in pro-
gress and costs are not complete. We omitted the B-1 program from
the data set for generating some flight test and ST&E CERs because
the B-1A’s flight-test program was unusually long, resulting in
atypical costs. Although the B-2’s values were considerably larger than
those for the other programs, our analysis included this aircraft
because it provides the only example of both a stealthy aircraft and a
continuous development effort for a modern bomber. As summarized
in Table 5.5, CERs for five categories of test costs—total contractor
ST&E, ground test, static and fatigue tests (which are a subset of
ground test), flight test, and other test—are presented below.
Following that table is a series of subsections that present the
CERs we developed for each category with a short discussion of each.
Table 5.4
Aircraft Variables Correlation Matrix (11 Programs)
NR GND + AC PGM FLT

DEV STE GND FLT FLT OTHER ST/F T1 WE Mo DUR DUR HRS F/A CGO
NR DEV 100
STE 93 100
GND 92 97 100
FLT 63 80 64 100
GND + FLT 91 100 97 81 100
OTHER 95 99 95 77 97 100
ST/F 61 63 91 14 67 50 100
T1 96 95 98 58 94 95 94 100
WE 40 42 58 8 47 33 99 51 100
AC Mo 48 53 36 75 52 54 –35 36 –35 100
EMD DUR 74 85 78 81 86 81 52 73 47 42 100
DUR 54 76 67 82 77 71 40 58 35 42 94 100
FLT HRS 63 58 58 46 59 55 28 58 28 61 35 16 100
F/A –51 –52 –61 –24 –54 –47 –71 –58 –74 10 –73 –63 –11 100
CGO –11 –22 –3 –51 –19 –26 54 –3 58 –62 1 –7 –19 –67 100
Total Contractor ST&E

Figure 5.4 compares the actual and predicted costs of the preferred
CER. Table 5.6 is the preferred CER, which excludes the B-1A/B
program on the assumption that its cancellation and restart intro-
Table 5.5
Summary of T&E Estimating Resources
CER CER
(Preferred) (Alternative) Database
Contractor ST&E • • •
Ground tests • • •
Static and fatigue test • •
Other ground test •
Flight tests • • •
Other tests • •
Government DT&E •
Government OT&E •
Figure 5.4
Total Contractor ST&E Cost CER
Predicted contractor ST&E costs
Actual contractor ST&E costs

RAND MG109-5.4
duced atypical inefficiencies. Tables 5.7 and 5.8 present alternative

CERs. Table 5.7 does include the B-1. The preferred and alternative
CERs suffer from the limitation of requiring an estimate of non-
recuring development cost as an input.
Table 5.6
Total ST&E CER—Preferred
CER STE = 215.9 + 0.2757NR DEV

Standard error 357.11
F-statistic 78.91
t-statistic on independ-
ent variable 8.89
Coefficient of variation 34.62%
Adjusted r 2 89.64%
Number of observations 10
NOTE: Omits B-1.
Table 5.7
Total ST&E CER—Alternative 1
CER STE = 1.828(NR DEV)0.806

F-statistic 42.72
ent variable 6.54
Adjusted r 2 in unit space 83.75%
NOTE: Includes B-1.
Table 5.8
Total ST&E CER—Alternative 2
CER STE = 2.509(NR DEV)0.7586

F-statistic 35.36
ent variable 5.95
NOTE: Omits B-1.
Contractor Ground Testing

Figure 5.5 compares the actual and predicted contractor ground test
costs of the preferred CER, shown in Table 5.9. Table 5.10 presents
an alternative CER using nonrecurring development cost and empty
weight.
The T1 cost for the FSD/EMD units proved to be the best pre-
dictor of ground-test costs. Although weight was a good predictor for
static and fatigue tests, the recurring unit cost of the aircraft is a bet-
ter predictor when other ground-test costs, such as subsystem and
avionics integration, are included. The CER has the undesirable fea-
ture of requiring an estimated first unit cost to generate the estimate
of ground-test costs.
Figure 5.5
Contractor Ground-Test CER
Predicted contractor ground test costs
Actual contractor ground test costs

RAND MG109-5.5
Table 5.9
Ground-Test CER—Preferred
CER GND = 94.84 + 1.54T1

F-statistic 279.10096
ent variable 16.706
Coefficient of variation
based on standard
error (standard
error/avg act) 20.86%
Adjusted r 2 96.53%
Table 5.10
Ground-Test CER—Alternative
CER GND = (–13.32) + 0.1299NR DEV + 0.001278WE

F-statistic 39.21
t-statistic on first inde-
pendent variable 7.02
t-statistic on second
independent variable 2.13
Adjusted r 2 88.43%
NOTE: Empty weight is significant at 93 percent.
Static and Fatigue Testing

Static and fatigue test costs are a significant portion of ground-test
costs. Because these costs are classified and reported relatively consis-
tently, we were able to develop a single satisfactory CER (Figure 5.6
and Table 5.11).
Empty weight is the independent variable. Our data source for
the B-2 did not identify the cost of static and fatigue tests, so we
could not include it in the data set.
Figure 5.6
Static and Fatigue Test Costs
Predicted contractor static and fatigue
test costs
Actual contractor static and fatigue test costs

RAND MG109-5.6
Table 5.11
Static and Fatigue Test CER—Preferred
CER ST/F = 37.21 + 0.001573WE

F-statistic 263.68
ent variable 16.24
Coefficient of variation
based on standard
error (standard
error/avg act) 18.94%
Adjusted r 2 96.69%
NOTE: Omits B-2 because of the lack of data on static and fatigue costs.
Flight Testing
The CER in Figure 5.7 and Table 5.12 provided the best predictive
value for contractor flight-test costs, capturing both the fixed (dura-
tion) and variable (aircraft months) aspects of flight testing. Aircraft

months are defined as the total number of months particular aircraft
are assigned to the development flight-test program. Duration is
defined as the number of months from first flight to end of DT. The
above relationship has the undesirable feature of a large negative
intercept, which makes it especially important to use the CER within
the range of the historical data and to check the results against
selected analogies.
Alternative CERs are provided in Tables 5.13 and 5.14. The
second alternative CER (Table 5.14), which excludes the B-1A/B, has
a similar coefficient of variation and eliminates the negative intercept.
(Dropping the B-1 does not significantly change the coefficient of
variation in the linear formulation.)
Figure 5.7
Contractor Flight-Test CER
Predicted contractor flight-test costs
Actual contractor flight-test costs

RAND MG109-5.7
Table 5.12
Flight-Test CER—Preferred
CER FLT = (–311.7) + 1.736AC Mo + 5.268DUR

F-statistic 48.51
Adjusted r 2 91.35%
NOTE: Aircraft months were not available for the T-45.
Table 5.13
Flight-Test CER—Alternative 1
CER FLT = 0.01887(AC Mo)1.454(DUR)0.4879

F-statistic 33.76
NOTE: T-45 omitted due to lack of data on aircraft months. Duration is significant at
70 percent.
Table 5.14
Flight-Test CER—Alternative 2
CER FLT = 0.04654(AC Mo)1.475(DUR)0.2137

F-statistic 39.92
NOTE: T-45 omitted due to lack of data on aircraft months. B-1 omitted because of
stretch out. Duration is significant at 81 percent.
Another caveat is that these CERs estimate contractor flight-test

costs only. There is some evidence that the amount of government
cost and effort on flight-test programs is increasing over time. In fact,
government costs have been larger than contractor flight-test costs on
the two most recent fighter flight-test programs, the F/A-18E/F and
projections for the F-22.
Figure 5.8 shows the percentage of the flight-test programs that
government activities perform. The contractor and government
efforts were largely separate for the F-14 (first flight 1971) and the
F/A-18A/B (first flight 1978).6 The F/A-18E/F and F-22 programs
are from the 1990s and early 2000s, respectively. By this time, test
Figure 5.8
Percentage of Government Costs in Flight Test Have Been Increasing
100
90
80
Percentage of flight-test costs
70 65%
60
52%
50
40
30 27%
20 17%
10
0
F-14* F/A-18A/B* F/A-18E/F F-22**
* Total government costs shown.
** In progress.
RAND MG109-5.8
_____________
6 In these two cases, the government costs shown represent total program office spending on the pro-
gram during the flight-test period and are thus an upper bound on what the program office could have
spent on testing alone.
programs had become integrated contractor and government efforts,

and the programs were paying for most of the variable costs of test-
ing. We did not have government cost data on enough programs to
generate a CER that included government costs, but analysts with
access to the proprietary supplement can use the CERs for contractor
costs in conjunction with the government costs on individual pro-
grams along with information on the program being estimated to
estimate the total cost of flight test.
Other Contractor Test Costs

Figure 5.9 compares actual and predicted costs, and Table 5.15 pre-
sents the CER we developed for other test costs.
This category includes a variety of efforts that are not clearly
associated with either flight or ground tests, such as test support,
planning, and test requirements. The nature of the activities included
in this subtotal makes it logically related to the scope and complexity
Figure 5.9
CER for Other Contractor Test Costs
Predicted other contractor test costs
Actual other contractor test costs

RAND MG109-5.9
Table 5.15
Other Contractor Test CER—Preferred
CER Contractor Other Test = (–134.5) + 0.5041(Estimated

Contractor Ground + Flight Test)
F-statistic 225.9
ent variable 15.03
Adjusted r 2 95.74%
of the rest of the test program. The CER estimates other test costs as
a function of the estimated costs of ground and flight test.
A Priori Expectations for Guided-Weapon ST&E

Relationships
Our analysis of guided-weapon ST&E costs was more constrained
than was our analysis of aircraft test costs primarily because of the
variety of classes of systems in our data set. Table 5.16 summarizes
the characteristics of the data set. The table shows that the data set of
weapon programs includes weapons powered for medium range (air-
to-air or air-to-ground missiles) or long-range (cruise missiles) flight,
or unpowered; new development programs or modifications of
existing weapons; and test programs of varying length and scope.
We identified cost drivers for weapon ST&E from our discus-
sions with personnel in program offices and test organizations and
from our analysis of the data collected. Our identification of cost
drivers guided the development and selection of CERs, although we
were often unable to find a satisfactory quantitative expression of cost
relationships because of the familiar (for cost estimators) problems of
too many variables for too few observations.
The first cost driver considered for missiles and munitions was
the type of weapon. This set of cost drivers divided our data set into
three weapon classes. We expected and observed that unpowered
Table 5.16
Summary Characteristics of Missiles and Guided Munitions
Type of DT Guided
Type of Devel. or Live DT
Weapon Program Launches Months FSD/EMD Start
Phoenix Air-to-air
missile New 64 43 December 1962
AMRAAM Air-to-air
missile New 86 44 December 1981
AMRAAM Air-to-air
Ph 1 missile Mod 12 24 March 1991
AMRAAM Air-to-air
Ph 2 missile Mod 13 89 June 1994
AIM-9X Air-to-air
missile Mod 20 33 December 1996
AMRAAM Air-to-air
Ph 3 missile Mod 8 18 December 1998
HARM Air-to-
ground
missile New 23 18 August 1978
IIR Maverick Air-to-
ground
missile Mod 52 21 October 1978
Harpoon Cruise
missile New 33 15 June 1971
SLAM-ER Cruise
missile Mod 8 15 March 1995
JASSMa Cruise
missile New 10 19 November 1998
SFW Unpowered
munition New 25 39 November 1985
JSOW Baseline Unpowered
munition New 24 N/A June 1992
JDAM Unpowered
munition New 170 N/A October 1995
WCMD Unpowered
munition New 61 20 January 1997
a
In progress; not used for CER development.
munitions, such as JDAM, JSOW, and WCMD, were the least

expensive to test as a class of weapons. The complex sensor-fuzed
weapon (SFW) was relatively expensive for this group. We expected
and observed that cruise missiles were more expensive to test than
shorter-range missiles on a cost per shot basis.
The second cost driver was whether the program was a new
development or a modification of an existing system. Logically, we
expected new development programs to be more expensive to test.

Among the air-to-air missiles, for example, the AMRAAM pre-
planned product improvement (P3I) programs were less expensive to
test than the more-ambitious AIM-9X modification program or the
original AMRAAM FSD program. This second cost driver further
subdivided our data set because some programs within the air-to-air
and cruise missile types were new developments and some were modi-
fications.
A third cost driver is the complexity of the total system or mis-
sile component that is being developed and tested. Unlike the first
two cost drivers, complexity as it relates to testing is not easy to define
or to measure consistently. Indicators of complexity include the
number and types of sensors and operational modes. Among the
unguided munitions, SFW is complex because of its self-contained
sensors and the multiple stages the weapon goes through from launch
until it fires. It had the largest test costs in total and on a cost-per-
shot basis among its class. Among the air-to-air missile modification
programs, the AIM-9X made the most extensive changes, with newly
developed guidance and tail control systems. It had the highest test
costs among modification programs. Typical parameters for measur-
ing the complexity of the system, such as operational modes, weight,
and density, were not very useful in characterizing T&E for modifica-
tion programs, which is more a function of the nature and extent of
the modifications performed. To attempt to capture both the degree
of modification and system complexity, we used contractor develop-
ment cost less T&E.
Once weapon programs were categorized by type, new develop-
ment or modification, and complexity, we expected that the number
of flight tests would drive ST&E costs. Flights with launches are
more costly than captive-carry flights, and guided launches more
costly than separation launches.7 For example, we expected to find a
relationship between ST&E costs and the number of captive-carry
flights and live shots for air-to-air missiles. These programs typically
_____________
7 It should be noted that certain complex captive-carry flight tests can be nearly as expensive as guided
launches.
require a great deal of captive-flight testing to assess the fidelity of

simulation models, provide sensor data for the models and software
development, test missile aerodynamic and integration issues with the
platform, etc. We expected that the relationship between flight tests
and test costs for the unpowered weapons that are also part of the
data set would be weaker. In general, we expected the number of
launches to be a cost driver for each type of weapon. We also
expected that the cost per shot would be very different for an unpow-
ered bomb kit with limited range, such as JDAM, and an expensive,
long range, cruise missile, such as JASSM or the Standoff Land-
Attack Missile–Expanded Response (SLAM-ER), even ignoring the
cost of the test article.
Results of Weapon ST&E Analysis

We began our efforts to develop CERs from the entire data set of
weapon programs. However, the cost behavior of the unpowered
weapons (JDAM, JSOW, WCMD, and SFW) could not be satisfac-
torily modeled with the available data. Dropping the unpowered
weapons from the data set resulted in a satisfactory CER for total
(contractor and government) guided-weapon test costs for missiles.
Table 5.17 defines the variables for missiles and contains the abbre-
viations used in Table 5.18. Table 5.18 is a correlation matrix.
The preferred CER for missiles (Figure 5.10 and Table 5.19)
estimates total DT costs as a function of contractor nontest develop-
ment costs and number of launches. This form accommodated both
new development and modification programs and reflected the vari-
able costs in the number of launches.
We developed an alternative CER (Table 5.20) using number of
launches and DT duration, dropping AMRAAM Phase 2 because of
its unusually long duration. The other candidate independent vari-
ables were not as robust across the range of systems.
Table 5.17
Variable Definitions for Missiles
Variable Name Variable Definition Units

CTR DEV (Contractor development less test) FY 2001 $M
Contractor cost for the missile
development programa minus the
cost of ST&E
CTR TEST (Contractor test) Contractor test costs FY 2001 $M
in; does not include government
costs of DT or OT
GOVT DT (Government DT) Government costs FY 2001 $M
for DT; does not include the cost of
OT
STE (ST&E) Total contractor and FY 2001 $M
government test costs, excluding
OT
LAUNCH (Launches) Guided-weapon launches Number
in DT, not including separation test
launches or captive carry flights
DUR (Duration) Elapsed time from first Months
flight or launch to end of DT
MOD (Modification) Dummy variable for 1 = modification
program type program
0 = new development
AA (Air-to-air) Dummy variable for 1 = air-to-air missile
missile type 0 = other missile
a
Many cost reports used for weapons in this study did not differentiate between non-
recurring and recurring costs.
Table 5.18
Missile Variables Correlation Matrix (10 Programs)
CTR CTR GOVT

DEV TEST DT STE LAUNCH DUR MOD AA
CTR DEV 100
CTR TEST 64 100
GOVT DT 84 31 100
STE 92 59 95 100
LAUNCH 77 50 90 93 100
DUR 14 –17 27 17 13 100
MOD –79 –44 –67 –71 –63 8 100
AA 23 –11 29 21 9 56 17 100
Figure 5.10
Missile Total ST&E CER (Excludes Guided Munitions)
Predicted missile ST&E costs
Actual missile ST&E costs

RAND MG109-5.10
Table 5.19
Missile Total ST&E CER—Preferred
CER STE = 15.34 + 0.07078CTR DEV + 1.401LAUNCH

F-statistic 103.82
Adjusted r 2 95.81%
NOTE: Includes missile programs only. The equations and the independent variables
are significant at the 98 percent level of confidence or higher except where noted.
Table 5.20
Missile Total ST&E CER—Alternative
CER STE = (–18.35) + 1.713LAUNCH + 2.244DUR

F-statistic 30.21
Adjusted r 2 87.96%
NOTE: Omits AMRAAM Phase 2; duration is an outlier. The duration variable is sig-
nificant at 90 percent.
Developing a T&E Cost Estimate

As discussed above, the estimating sections of this report were written
primarily for cost analysts who are tasked either with developing
estimates of test programs before detailed test planning data are avail-
able or with developing independent estimates and assessments of
program office T&E estimates.
The first step in developing an estimate for a test program is to
determine its likely scope. Table 5.21 provides some indicators that,
in general terms, distinguish relatively complex test programs from
simpler efforts.
The primary missions of an aircraft or guided weapon are gen-
eral indicators of its complexity. Fighter aircraft and air-to-air or
cruise missiles are more difficult to develop and test than trainers or
guided munitions. However, within broad missions, there can be
varying levels of expected performance, which will affect the amount
of development and therefore the level of testing involved. The test-
ing complexity for a fighter aircraft—achieving maximum levels of
speed, maneuverability, integration of onboard and offboard systems,
and signature reduction—can be significantly greater than for a pro-
gram whose performance requirements are less challenging.
In addition to mission- and performance-driven complexity, the
amount of new development in mission critical systems and subsys-
tems directly affects test programs. The programs we examined for

this study provide examples of high new-development content (F-22,
B-2, V-22, AMRAAM, SFW) and of more-evolutionary systems
(F-18E/F, B-1B CMUP, SLAM-ER, AMRAAM P3I). The system’s
Cost Analysis Requirements Description (CARD) or other program
documentation usually indicates which subsystems or key compo-
nents have been proven in other applications.
The planned activities and their schedule constitute a third class
of indicators of test program content. A general outline of the
required test program should be available early in the acquisition
process. Such test requirements as static and fatigue, full-scale live-
fire, extensive avionics and software, interoperability, LO, or a full
range of sensor and seeker testing are key indicators of test scope and
risk. For aircraft, developing a new engine is a major program in
itself, with corresponding test requirements.8 The planned duration
of the overall test program is another indication of both content and
risk.
Table 5.21
Indicators of Test Program Scope
Indicator More-Complex Examples Less-Complex Examples

Mission Air superiority Air transport
Air-to-air Munition kit
System Maximum performance Limited capability
Extensive integration Stand-alone
LO Conventional signature
Extensive functionality Single purpose
New technology Proven design
Design New design Limited enhancement
Test requirements Unusual test environments Typical environments
Compressed schedule Available slack
Multiservice Single service
Unusual security Typical security
Test planning is a continual balance of likely cost and risk.

Many aspects of testing have costs that are more a function of time
_____________
8 This monograph addresses only the integration testing of the engine and aircraft.
than of activity. Test schedules may be compressed in an effort to

reduce both testing and development costs. If this incentivizes higher
productivity and innovation on the part of government and con-
tractor testers, it can be an effective approach. However, arbitrary
schedule compression, especially when there are delays outside the
control of the test community, such as late delivery of testable hard-
ware or delays and retesting for correction of defects, commonly
results in slipped schedules and increased costs.
After determining the likely scope of the test program, the next
step is to compare the planned program with similar historical cases.
As discussed throughout this monograph, T&E processes and
approaches continue to evolve. However the realism of program plans
that project large T&E savings must be carefully evaluated in view of
historical experience. Appendix A briefly describes the technical, pro-
grammatic and T&E aspects of eight aircraft and eight guided-
weapon programs. Information on some older programs may be
found in several studies listed in the References. We have attempted
to provide enough information to allow the analyst to draw meaning-
ful comparisons of test program scope, schedule, and cost. The corre-
sponding cost data are contained in a limited-access supplement to
this report. Using these resources, programs of similar content can be
identified.
As a first step, an estimate of total contractor T&E cost for air-
craft programs can be made using the parametric CERs provided
above. If sufficient test program information is available, this estimate
can be refined considerably by using the individual CERs for ground,
flight and other test costs. Estimates of government costs can be
developed by analogy with similar programs from the data supple-
ment. For missile programs, the preferred and alternative CERs can
be used to estimate both the contractor and government portions of
the DT program.
This estimate should then be compared to the actual costs and
durations of similar programs in the database. In the best case, several
programs may be similar in key aspects to the program being esti-
mated. Rudimentary data analysis will provide averages and ranges for
total test costs and durations, highlighting outliers. When more-
detailed data are given, major test activities can also be compared
directly. When adequate data are available, regression analysis can be
used to develop tailored CERs at the total level or for selected por-
tions of the test program. If sufficient comparable data are not avail-
able to derive meaningful statistical relationships, it may be useful to
use averages and ranges for several analogous programs or values from
a single comparable program, adjusted for differences in the program
being estimated.
As always when dealing with small data sets, cross checks should
be used to confirm estimates. Alternative metrics that can be useful to
develop cross checks include the following:
• cost per flight hour (aircraft)
• cost per shot (guided weapons)
• cost per test month.
CHAPTER SIX
Conclusions and Recommendations
The following conclusions can be drawn from the data we collected

and our interviews with government and contractor test personnel:
• The overall cost of T&E has been a relatively constant propor-
tion of aircraft and guided-weapon system development costs
since the early 1970s. Despite increasing use of M&S,
improvements in instrumentation and test processes, reduction
of dedicated government testing, and various acquisition
streamlining initiatives, T&E costs have remained relatively con-
sistent. Although various explanations of this situation are possi-
ble, the dominant factors are probably the increasing complexity
of the systems being tested and greater test program content.
• M&S is now integral to most test programs. In fact, in several
cases, M&S capabilities were hard pressed to meet the program’s
T&E requirements. In many programs, the analytical tools were
not always adequate to confidently waive live testing, but in all
cases, M&S was judged to be a good investment that at least
reduced the risk, and often the duration, of live tests. In addi-
tion to direct benefits for T&E, robust M&S has other benefits
for
– evaluating design excursions during development
– developing tactics
– training operators
– evaluating future system enhancements.
107
• Although there is some disagreement about the appropriate level

of testing in specific circumstances (e.g., live-fire testing, testing
for statistically rare events), we found little controversy in gen-
eral over the scope of testing. Several sources, however, expressed
the opinion that thoughtful reevaluation of test procedures
could improve the pace and efficiency of the typical test pro-
gram.
• There was general agreement that integrated contractor-govern-
ment test teams were a positive force for optimizing testing.
Similarly, combined development and operational test teams
were judged to have been valuable because they avoid redundant
testing and identify operational effectiveness and suitability
issues for early resolution. Some program personnel expressed a
desire for even more intensive “early involvement” from the
operational test community. The primary constraint appears to
be the limited staffing of the services’ operational test organiza-
tions.
• It is too early to assess the outcome of some recent innovative
test management approaches that involve giving the contractor
broad latitude in developing and executing the DT program.
Another innovative approach—relying on non-DoD tests and
certifications of nondevelopmental aircraft for DoD applica-
tions—has not been as successful as hoped. In the two cases we
examined (both trainer aircraft), the DoD requirements were
different enough from those of previous customers to require
significant modification and testing. In both cases, the problems
were more the result of underestimating the differences in
requirements and the scope of required modifications than of
quality problems with previous testing and certification proc-
esses.
• Data on costs incurred by government organizations are much
more difficult to collect and document than are the corre-
sponding contractor cost data. This did not initially seem to be a
serious limitation because we assumed that acquisition reform
would lead to decreasing government costs as contractors took
on a greater share of the effort. For programs where we were able
Conclusions and Recommendations 109
to obtain government costs, this was not generally the case. In

many instances, contractors still rely on government test facili-
ties and functional expertise, particularly for high-cost, low-
utilization test capabilities. Government personnel normally par-
ticipate actively in the integrated test teams. Even in programs
that do not constrain the contractor’s choice of test facilities, the
government facilities can find themselves acting as test subcon-
tractors to the system prime contractor. Of course, most open-
air testing continues to take place on DoD ranges. Our
recommendation is that government cost data be consistently
accumulated and reported, just as contractor data is today. This
would ensure that the program’s total financial picture would be
available for management in the present and analysis in the
future. This would help government test facilities better evaluate
the cost and schedule implications of their processes, allowing
them to better assess the contribution of all their activities and
to focus investment and management attention on those
deemed most critical to their customer base.
APPENDIX A
Aircraft Program Descriptions
B-1B Conventional Mission Upgrade Program Block D
Mission
The B-1B is a long-range supersonic bomber, originally designed to
penetrate sophisticated air defenses. 1 The CMUP enhances the
B-1B’s capability to deliver modern conventional weapons. Modi-
fications are being incorporated in four combined hardware and
software block upgrades. This description summarizes only the
CMUP Block D upgrade test program.
System Description
The Block D upgrade to the B-1B consists of the following hardware
and software modifications:
• Global Positioning System (GPS) capability for navigation,
offensive avionics, and weapon delivery
• jam-resistant very high and ultrahigh frequency radios
• MIL-STD-1760 interface to the Multipurpose Rotary Launcher
to accommodate advanced conventional munitions
• JDAM integration
• offensive system and ancillary software sustainment upgrades
_____________
1 Information on the B-1B came from SPO input, Air Force Flight Test Center (2000) for
DT/IOT&E, and AFOTEC (1998) for dedicated OT&E.
111
• defensive system software improvements.
Programmatics
Boeing North American, the system prime contractor, was awarded
the contract to integrate the B-1 CMUP modifications. Program-
matic milestones included the following:
• EMD contract award: March 1995 (Contract F33657-94-C-
0001)
• critical design review (CDR): May 1996
• production approval: July 1997
• production contract award: July 1997 (Contract F33657-97-C-
0004).
Test Program
The combined DT&E/IOT&E focused on verifying the GPS
incorporation, sustainment software upgrades, reincorporation of
ground moving target indication and tracking, defensive system soft-
ware upgrades, and JDAM integration.
Initial avionics testing was conducted at the contractor’s avionics
and system integration laboratories, the Avionics Integrated Support
Facility at Tinker AFB, and the Integrated Facility for Avionics Sys-
tem Test at Edwards AFB. The avionics software modifications were
accomplished in 18 months because software anomalies were identi-
fied during ground and flight tests.
Initial JDAM testing took place at Arnold Engineering Devel-
opment Center, followed by ground testing; captive-carriage, safe-
separation, environmental conditions, safety, and electromagnetic
interference (EMI) and electromagnetic compatibility (EMC) testing;
and actual flight tests at Edwards AFB. Approximately six JDAM
separation test vehicles and 23 guided test vehicles were dropped. In
addition to testing the performance of the Block D changes, the test
program also verified correction of previously identified system-level
deficiencies. Full-up live-fire testing of the B-1B was waived, but
component-level vulnerability testing was performed. There were
5,411 hours of major ground testing.
Aircraft Program Descriptions 113
Table A.1 summarizes the aircraft sorties made for the various
types of testing. Table A.2 provides data for the participating aircraft.
The three aircraft put in a total of 34 aircraft months of service.
AFOTEC conducted an operational assessment, combined
DT&E/IOT&E, and a one-month dedicated IOT&E. Approxi-
mately nine sorties each were used for combined testing and dedi-
cated IOT&E.
EMD began in January 1995, with the first test flight in March
1997. DT ended in September 1998; two aircraft made 97 flights,
flying for 660 hours. On average, the flight-test program achieved
20.6 flight hours per aircraft month, using engine-running crew
changes to maximize aircraft utilization.
Dedicated OT&E began in August 1998 and ended in Septem-
ber 1998. The three aircraft mentioned in Table A.2 provided six air-
Table A.1
B-1B Flight Testing
Test Flight
Aircraft Sorties Time
Event (no.) (no.) (hrs.)
Total EMD flight test

August 1997–September 1998 3 106 712
DT&E/IOT&E
August 1997–July 1998 2 97 660
Dedicated IOT&E
August 1998–September 1998 2 9 52
Table A.2
The B-1B Test Aircraft
Test Entered Exited Aircraft

Aircraft Testing Testing Months
85-0068 March 1997 September 1998 18

84-0049 July 1997 September 1998 14
85-0082a August 1998 September 1998 2
a
Used for dedicated OT&E.
craft months and flew a total of nine flights and 53 flight hours. Air-
craft 85-0082 was the Block D Kit Proof aircraft.
B-2A Spirit
Mission
The B-2 bomber combines LO, large payload, and long range to
deliver conventional or nuclear munitions. 2 The B-2 program began
during the early 1980s with the objective of penetrating sophisticated
air defenses and attacking high-value and heavily defended targets.
The design was modified for low-altitude operations during FSD.
Additional design and test efforts have been required to integrate pre-
cision conventional munitions, including the GPS-Aided Targeting
System, GAM, and JDAM.
System Description
The B-2 has a two-person crew and is powered by four General Elec-
tric F-118-GE-100 engines rated at 17,300 pounds of thrust. The
significant features include the following:
• reduced electromagnetic, infrared (IR), acoustic, visual, and
radar signatures
• extensive use of composite structures in the airframe
• fabrication, assembly, and finishing of parts to high tolerances to
achieve stealth
• a blended flying-wing shape
• two internal weapon bays
• 44,000-pound payload
• designed to carry general purpose bombs from 500 to 4,700
pounds, mines, GAMs, JDAM, and nuclear bombs3
• engine inlets and exhaust shaped for radar and IR stealth
_____________
2 Information on the B -2 came from CTF Highlight Summary Flight Test Production (undated, after
June 1997); T&E Master Plan for the Sustainment Phase of the B-2A Advanced Technology Bomber,
March 23, 1999; briefing slides from and discussions with the B-2 SPO.
3 A JSOW capability was added post-baseline.
• features offensive avionics, including Hughes Aircraft Com-

pany’s APQ-181 radar (now Raytheon’s)
• automatic terrain following to 200 feet.
Programmatics
A cost-plus-incentive-fee FSD contract was awarded to Northrop in
November 1981. This contract specified delivery of two ground-test
articles and six flight-test vehicles. Boeing and Vought teamed with
Northrop to design and build the airframe. Hughes Aircraft Com-
pany was the radar contractor, and General Electric Aircraft Engine
Group developed and built the engine. After flight testing, the EMD
aircraft were updated to the baseline (Block 30) configuration for use
as operational assets. There was no full-rate production. Program-
matic milestones included the following:
• aircraft preliminary design review: November 1982 and May
19844
• CDR: December 1985
• first engine delivery: December 1986
• LRIP contract: November 1987
• first flight: July 1989
• delivery of the durability test article: September 1989
• delivery of the static test article: January 1990
• first production aircraft delivery: December 1993
• end of EMD (baseline): March 1998.
Test Program
FSD/EMD began in November 1981, with first flight in July 1989
and DT ending in March 1998. The six test aircraft made 1,013
flights, flying 5,197 hours for a total of 310 aircraft months.
Dedicated IOT&E began in October 1993 and ended in June
1997. The six aircraft made 11 flights, flying 94.7 hours for these
tests.
_____________
4 The second review was for the low-altitude redesign.
Combined DT and IOT&E began in July 1989 and was com-

pleted in March 1998 (see Tables A.3 and A.4). A flying test bed was
used for early testing of radar and navigation systems. Test personnel
were organized as a CTF, which peaked at 2,011 personnel.
The B-2 did not fully meet operational requirements during the
baseline program. Deficiencies were found in the defensive manage-
ment, mission planning, and terrain following systems and in LO
maintainability, which affected maintenance man-hours per flying
hour and the mission-capable and sortie-generation rates. These issues
Table A.3
B-2 Flight-Test Program
Test Flight
Event Dates Aircraft Sorties Hours
Total FSD/EMD flight

testa July 1989–March 1998 6 1,013 5,197.0
Flying test bed b April 1987–December 1995 1 600 3,353.0
Dedicated IOT&E October 1993–June 1997 6 11 94.7
a
These tests accomplished 19,897 flight-test points and 4,194 ground-test points.
b
Avionics testing.
Table A.4
The B-2 Test Aircraft
Test Finish Build Aircraft

Aircraft Enter Testing Exit Testing Months Months
AV-1 July 1989 March 1993 3 41

AV-2 October 1990 August 1995 0 58
AV-3 June 1991 March 1998 12 69
AV-4 April 1992 June 1997 0 62
AV-5 October 1992 November 1997 7 54
AV-6 March 1993 December 1995 7 26
Total 310
NOTE: The Finish Build Months column shows the months that were spent purely on
finishing the building of the aircraft during the testing span. This work was necessary
for testing but was known to be incomplete at entry into testing. This time is not
counted in the Aircraft Months column. Also not included is the time spent installing
upgraded configurations resulting from testing.
were or are being addressed in the sustainment phase, along with

various upgrades. DOT&E waived requirements for full-scale live-fire
testing because the program was not expected to proceed beyond low-
rate production.
DT Ground Tests: Wind-tunnel and weapon-separation tests
took place at Arnold Engineering Development Center (AEDC).
Testing of the engine inlets was at a Northrop radio frequency range
in California. A six-degrees-of-freedom model for weapon separation
was used to simulate weapon drops (actual drops were made to verify
the model). The fatigue article at Palmdale, California, was tested to
two lifetimes. The static article was tested to 160 percent of ultimate
load.
DT Flight Tests: Until 1988, the flight-test plan called for four
test aircraft at 50 hours per month per vehicle. This plan was based
on a planned first flight in 1987, which in fact slipped to 1989. The
first six aircraft were delivered late and were incomplete, so the flight-
test plan was restructured in 1988 to six test vehicles at 20 hours per
month per vehicle. The first two flight-test aircraft (AV-1, AV-2)
were used for air vehicle testing and did not have the full avionics
suite. In-flight signature testing was done at a government facility.
Several unanticipated events during FSD/EMD affected the test
program including the following:
• a major redesign and retest of aircraft components early in FSD
to meet radar cross section (RCS) requirements
• a change in the primary mission from delivery of nuclear to con-
ventional weapons
• late delivery of partially complete aircraft.
IOT&E was followed by FOT&E Phase I conducted by
AFOTEC at Whiteman AFB through December 1998. FOT&E
testing has focused on correction of deficiencies.
C-17 Globemaster
Mission
The C-17 provides worldwide airlift for U.S. and allied combat
forces, equipment, and supplies.5 It can deliver passengers and out-
size, oversize, or bulk cargo over intercontinental distances without
refueling. The aircraft can land at conventional or austere airfields or
can make deliveries by airdrop. The capability for rapid, in-flight
reconfiguration allows the C-17 to transition easily among its mission
modes.
System Description
The C-17 is a four-engine turbofan transport aircraft. Its engines are
modified versions of an earlier commercial airline engine (PW-2040).
Significant features include the following
• A supercritical wing design and winglets reduce drag and
increase fuel efficiency and range.
• The aircraft can be refueled in flight.
• An externally blown flap configuration, direct lift-control spoil-
ers, and a high-impact landing gear system allow the aircraft to
use small, austere airfields.
• A forward and upward thrust-reverser system provides backup
capability, reduces the aircraft’s ramp-space requirements, and
minimizes interference with ground operations.
• The airdrop system is fully automated.
• A single loadmaster can operate the cargo door, ramp, and cargo
restraint systems, and off-loading equipment does not require
special handling equipment.
• The flight-control system is electronic (quad-redundant, fly-by-
wire).
• The two-person cockpit has multifunction displays.
_____________
5 Information on the C-17 came from T&E Master Plan for the C-17 Weapon System, dated August
1999; briefing slides and discussions with C-17 SPO; C-17 Program Office Estimate, dated June 1993;
C-17 Flight Test Progress Report for Month Ending December 31, 1994.
• The mission computer integrates the majority of the avionics.

• The Onboard Inert Gas Generating System handles fuel tank
inerting.
• Built-in test features reduce maintenance and troubleshooting
times.
Programmatics
A fixed-price-incentive-fee full-scale development contract with two
production options was awarded to McDonnell Douglas Aircraft
Company in December 1985. The contract was restructured in Janu-
ary 1988. LRIP was approved in January 1989. The lot III produc-
tion contract was awarded in July 1991, and the first flight was in
September 1991.
Test Program
Several developmental problems affected the test program. Software
integration was more complex than originally anticipated and
increased the amount of test effort, particularly for regression testing.
In an attempt to identify and solve problems on the ground,
McDonnell Douglas (now Boeing) established an avionics integration
laboratory and a flight hardware simulator in Long Beach, California.
Tables A.5 and A.6 summarize the flight-test program and aircraft
used, respectively.
FSD/EMD began in December 1985; first flight was in Sep-
tember 1991, and DT ended in December 1994. Six test aircraft flew
a total of 1,134 sorties and 4,096 flight hours. The test program used
a total of 169 aircraft months. Although the original plan had been
for 69 aircraft months and 2,277 flight hours (33.0 flight hours per
aircraft per month), the realized rate was 24.3 flight hours per aircraft
per month.
During static testing, the wing broke at 136 percent load and
had to be redesigned and retrofitted to production aircraft. The
original design included a hydromechanical flight-control system, but
wind-tunnel testing identified problems with it. As a result the sys-
Table A.5
C-17 Flight-Test Program
Test Flight % of
Aircraft Sorties Time Flight
Event (no.) (no.) (hrs.) Time
FSD flight testing

September 1991–December 1994
Planned testing
Avionics 565
Flying qualities 460
Aerodynamic performance 373
Aircraft systems 319
Mission systems 256
Structures 213
Test unique 44
Subtotal planned 2,230 54.4
Unplanned testing
Unplanned demand 1,074 26.2
Test and work requests 763 18.6
Added tests 29 7.0
Subtotal unplanned 1,866 45.6
Total FSD flight testing 6 1,134 4,096
DT/IOT&E
June 1992–December 1994 2
Dedicated IOT&E
December 1994–June 1995 2
NOTE: Total FSD testing included 5,623 flight-test points and 1,028 ground-test points.
Table A.6
The C-17 Aircraft Tested
Test Aircraft
Aircraft Entered Testing Exited Testing Months
T1 September 1991 December 1994a 39

P1 May 1992 December 1994a 31
P2 June 1992 October 1994 28
P3 September 1992 December 1994 27
P4 December 1992 December 1994a 24
P5 January 1993 September 1994 20
Total 169
a
At the end of DT.
tem, was changed to a four-channel fly-by-wire system. A mechanical

backup was added later, although the testing approach presumed total
reliance on the electronic system. Effective use of M&S allowed live
shots on a wing section to satisfy LFT&E requirements.
OT included AFOTEC’s early operational assessment, in Sep-
tember 1988, to support the Milestone IIIA decision, and an opera-
tional assessment in January 1990. AFOTEC determined that the
major risks for system development and IOT&E were software devel-
opment and avionics integration. Combined DT&E/IOT&E began
in June 1992. In support of a congressionally directed assessment of
the C-17, AFOTEC assessed the C-17 system as adequate overall,
considering its stage of development at the time, but identified range
and payload, maintainability, and software maturity as risk areas.
Dedicated IOT&E began in December 1994 and ended in June
1995 and involved two aircraft. The original plan was to dedicate
three aircraft months to training and eight aircraft months to
IOT&E. This testing was conducted in three phases. Phase I,
December 1–17, 1994, focused on C-17 cargo loading and transport-
ability. Phase II evaluated all operations except static line paratroop
drops and a “slice” of brigade airdrop demonstration. These areas
were evaluated in Phase III, which was completed in June 1995.
AFOTEC’s final IOT&E report included data generated during
initial squadron operations and the reliability, maintainability, and
availability evaluation. AFOTEC judged the C-17 to be operationally
effective and suitable, meeting all key parameters and demonstrating
outstanding direct-delivery capability and maintainability. Areas for
improvement included formation personnel airdrop, mission com-
puter takeoff and landing data, aeromedical evacuation capability,
fault isolation, support equipment, and software maturity.
F/A-18E/F Super Hornet
Missions
The F/A-18E/F is a carrier-based multimission strike fighter derived
from the F/A-18C/D. The F/A-18E/F’s range, payload, and surviv-
ability have improved over those of its predecessor.6 Missions include

fighter escort, combat air patrol, interdiction, and close air support.
System Description
The F/A-18E is a single-seat and F/A-18F is a two-seat combat-
capable trainer. Every part of the F/A-18E/F structure was redesigned
from its predecessor. In general, the structure was enlarged and
strengthened, the part count was reduced, and the use of materials
and tooling were changed. Ninety-six percent of the airframe unit
weight is unique to the E/F. The structure’s material composition
features more titanium and composites and less aluminum than its
predecessor. Changes from the F/A-18C/D include the following:
• The gross landing weight increased by 10,000 pounds.
• Redesigning the forward fuselage increased its strength and
decreased the part count.
• The center-aft fuselage was lengthened 34 inches to increase fuel
capacity.
• Wing area increased 100 ft2 , and the wingspan increased by
more than 4 ft, also increasing internal fuel capacity.
• The areas of the control surfaces, horizontal tail surfaces, and
leading-edge extension increased.
• Unitizing reduced the part count by 42 percent from that of the
C/D and reduced manufacturing costs.
• The new configuration can carry an additional 3,600 pounds of
fuel internally and 3,100 pounds externally.
• The aircraft has two additional hard points for weapons.
• Incorporating such low-cost stealth features as saw-toothed
doors and panels, realigned joints and edges, and angled anten-
nas reduced the RCS.
_____________
6 The test program description came from TEMP No. 0201-04 Rev. B; the dates, number of flights,
and aircraft months came from the EMD flight log that the program office provided; the system descrip-
tion comes from various sources.
• Although the E/F and the F/A-18C/D have many common sub-
systems, the E/F has more powerful actuators to accommodate
larger control surfaces.
• Over 90 percent of the avionics are common with the F/A-18C.
• General Electric developed the F414-GE-400 turbofan engine,
rated at approximately 22,000 pounds thrust, for the new air-
craft. Two engines power the E/F.
Programmatics
The Navy awarded a cost-plus-incentive-fee contract for airframe
EMD to McDonnell Douglas (now Boeing) in June 1992. Northrop
is the major airframe subcontractor and is responsible for the center
and aft fuselage, vertical tail, and several subsystems. Three ground-
test articles—static, drop, and fatigue—were built, and seven flight-
test vehicles were built and flight tested in EMD.
• EMD contract award: June 1992
• CDR: June 1994
• first flight: November 1995
• fatigue testing completed: July 1998
• flight testing completed: April 1999.
Test Program
Two phases of studies and testing preceded EMD. The configuration
study phase (1988 to 1991) used approximately 600 hours of wind-
tunnel tests of a 12-percent scale model to define the current aero-
dynamic configuration and engine performance requirements. Dur-
ing the pre-EMD phase (1991 and 1992), an additional 2,000 hours
of wind-tunnel tests on models helped optimize the configuration
and reduce design risks before EMD.
DT in EMD was conducted in several phases. During DT-IIA
(November 1995 to November 1996), the static test article was used
for initial wing-bending tests; the drop-test article was used for a
series of landing-gear tests at increasing sink rates; the manned flight
simulator was used for aircrew training and other tests; and flight
testing focused on envelope expansion.
FSD/EMD began in June 1992 and ended in April 1999, with

first flight in November 1995. Eight aircraft made 3,141 flights, log-
ging 4,620 flying hours, for a total of 244 aircraft months. OPEVAL
involved seven aircraft and began in May 1999 and ended in
November 1999. Tables A.7 and A.8 summarize the flight-test
program and aircraft used, respectively.
Table A.7
F/A-18E/F Flight-Test Program
Test Aircraft Sorties Flight Time

Event (no.) (no.) (hrs.)
Total FSD Flight Test

November 1995–April 1999 8 3,141 4,620
Avionics 600
Flying qualities 1,890
Prop. performance 195
Aircraft systems 275
Armament 310
Structures 1,126
Carrier suitability 189
Other 35
OPEVAL
May 1999–November 1999 7
NOTE: One LRIP aircraft was used for two months in FSD testing.
Table A.8
The F/A-18E/F Test Aircraft
Test Aircraft
Aircraft Enter Testing Exit Testing Months
E-1 November 1995 April 1999 41

E-2 December 1995 April 1999 40
E-3 January 1997 April 1999 27
E-4 July 1996 April 1999 34
E-5 August 1996 April 1999 32
F-1 April 1996 April 1999 37
F-2 October 1996 April 1999 31
F-4 February 1999 April 1999 2
Total 244
In March 1996, early in the flight-test program, there was a

wing drop incident. 7 This occurred during high-speed maneuvers and
prevented the pilot from performing close-in tracking maneuvers on
potential adversaries. After identifying the wing-drop problem, a
Boeing-Navy team performed wind-tunnel tests and computational
fluid-dynamics studies to identify the cause. The results indicated
that the problem was associated with airflow separation differences
between the left and right wings. Boeing and the Navy considered
three solutions to the problem and implemented a change to the wing
to correct the problem.
During DT-IIB (December 1996 to November 1997), static
and drop testing were completed; fatigue testing began; and flight
testing focused on expanding the flight envelope, initial sea trials,
evaluation of the aircraft in the carrier environment, evaluation of
aircraft subsystem performance, and start of EW suite and IR signa-
ture testing.
In DT-IIC (December 1997 to November 1998), engine full
production qualification ground tests were completed during over
10,000 hours of testing. LFT&E was conducted using the drop-test
article. Live-fire tests included analysis of previous F/A-18A/B live-
fire tests and used a ground-test article for eight major live-fire tests to
evaluate vulnerability of the F/A-18E/F tail, wing, and fuselage. IR
signature and EW suite tests were completed. Flight testing included
dynamic RCS measurements, flight-envelope expansion tests, and
weapon clearance tests.
DT-IID (November 1998 to April 1999) was the TECHEVAL.
Testing focused on validation and verification of production-
representative weapon software functionality, EW suite testing, and
testing in the carrier environment.
DT flight testing was finished at the end of April 1999, com-
pleting over 15,000 test points and clearing 29 weapon configura-
tions for flight.
_____________
7 Wing drop is an abrupt, uncommanded rolling motion of the aircraft during certain flight conditions.
OPEVAL (May 1999 to November 1999) involved seven pro-

duction aircraft. Much of the testing was at China Lake, California,
with deployments to other locations, including carrier operations.
F/A-22 Raptor
Mission
The F/A-22’s primary mission is air superiority, with a secondary air-
to-ground mission when equipped with JDAM.8 Its combination of
sensors, displays, weapons, and LO is designed to provide first-look,
first-kill capability in all tactical environments. It will eventually
replace the F-15 in the air superiority role. Its designation was
changed from F-22 to F/A-22 in September 2002 to recognize its
dual role.
System Description
The F/A-22 Raptor is a twin-engine, single-seat, LO, all-weather
fighter and attack aircraft. The following are some of its key features:
• Advanced turbofan engines, which allow sustained supersonic
cruise without afterburners, and thrust vectoring provide
enhanced performance.
• Reduced radar and IR signatures and internal weapon carriage
provide LO.
• The advanced integrated avionics include
– fusion of radar, EW, and communications, navigation, and
identification sensor outputs
– long-range, active and passive, offensive and defensive sensors
to improve detection and tracking
_____________
8 Information on the F-22 came from the F-22 T&E Master Plan, Version 1.0, dated September 2000;
Draft F-22 CARD, dated April 1999; briefing slides and discussions with F-22 SPO; From the
Advanced Tactical Fighter to the F-22 Raptor, ANSER, March 24, 1998.
– multispectral, wide-aspect threat warning and tailored multi-

spectral defensive response with automated and manual
modes
– modular open architecture with inherent growth capability.
The operational software, which is developed and released in
blocks, consists of approximately 2.1 million lines of code.
• Automation and optimized pilot interfaces provide expanded
situational awareness.
• Onboard support systems with extensive integrated diagnostics
and fault and failure tolerance capability provide improved sup-
portability; accessibility has been simplified by reducing the tool
set and the amount of unique support equipment.
Programmatics
The Advanced Tactical Fighter entered its DEM/VAL phase in
October 1986 and completed it in August 1991. This phase pro-
duced the YF-22 (Lockheed, Boeing, and General Dynamics) and
YF-23 (Northrop and McDonnell Douglas) prototypes to demon-
strate airframe and engine design approaches.
In August 1991, the F-22 entered EMD, with Lockheed Martin
selected as the system contractor and Pratt & Whitney as the engine
contractor. The EMD contract specifies nine EMD aircraft, two
ground-test articles, and a full scale pole model for RCS testing. In
1996, the Air Force Acquisition Executive, concerned about cost
growth trends on the program, chartered a joint estimating team
(JET) to assess the F-22’s costs and schedule. The JET recommended
delaying the transition to production and adding 12 months for
completing avionics development. The restructured program dropped
the preproduction verification aircraft. Congress imposed a cost cap
on development and production that was based on the JET
projections, as well as exit criteria the program had to meet before
Congress would authorize the transition to production.
The EMD aircraft first flew in September 1997. In May 1998, a
contract for two production-representative test vehicles (PRTVs) and
a first lot of six production aircraft was awarded. A separate “program
support” contract for contractor tasks not directly identifiable to a
specific aircraft was let to provide cost traceability to the negotiated

target price curve. In its FY 2000 appropriations bill, Congress
approved redesignating the six lot I aircraft as PRTV lot II and pro-
curing them under RDT&E funding. As of October 2002, all nine
test aircraft had been delivered, and the test aircraft to be used in
dedicated IOT&E were undergoing structural and electrical system
modifications.
Test Program
The F/A-22 test program is representative of the most complex pro-
duction aircraft testing because it combines a new, advanced-design
airframe, engine, avionics, and LO features in a single, highly inte-
grated system. This translates into additional testing complexity (see
Table A.9 for a summary).
The test programs during DEM/VAL focused on allowing the
two competing contractor teams to demonstrate key technologies and
risk reduction. It was not a competitive “fly-off.” The government
did not specify the testing, and the test results were not a deciding
factor in the EMD source selection. The contractors, with govern-
ment participation, executed a short but intense set of flight demon-
strations.
One carryover from the DEM/VAL phase is the use of a highly
modified Boeing 757 flying test bed to test and troubleshoot F/A-22
avionics and software before its installation on the F/A-22. The test
bed has an F/A-22 forward fuselage grafted onto its nose and a wing
for mounting sensors attached to the upper fuselage, above the cock-
pit.
The program also used the Variable Stability In-Flight Simula-
tor Test Aircraft, an F-16 specially configured to mimic the flying
characteristics of the F/A-22. This aircraft was used to verify the flight
control laws to be used in the F/A-22 flight control system.
The F/A-22 CTF consists of DT&E and OT&E test personnel
from the contractors, the Air Force Flight Test Center, AFOTEC,
the F/A-22 SPO, and Air Combat Command. Of the 770 personnel
on the CTF in late 2001, 480, or 62 percent, were government (mili-
Table A.9
Special Requirements for F/A-22 Testing
Feature Effects on Testing
Reduced signature design Requires

• maintaining several test aircraft in the LO
configuration
• obtaining and scheduling unique LO test assets
• managing security considerations.
Internal weapon carriage Requires
• additional wind-tunnel characterization of flow
field with bay open
• an additional flight-test configuration (doors
open) for performance and flying qualities.
Sustained supersonic cruise Reduces test time with chase aircraft.
Requires additional tanker support.
Increases use of supersonic test airspace.
Thrust vectoring Requires
• special ground-test fixtures to control vectored
exhaust gases
• multiaxis force and moment instrumentation for
thrust measurement
• ground and in-flight performance testing
• expanded flying and handling quality testing
• failure modes and effects testing, particularly
with respect to asymmetric actuation.
Integrated avionics Requires
• additional EMI and EMC testing
• comprehensive ground and air testing of
integrated system modes
• collecting data on system timelines and their
effect on system performance.
Sensor fusion Requires
• high-density, multispectral, integrated, and
enhanced-fidelity target and threat simulation
• comprehensive integrated ground-test facilities.
Highly integrated wide-field- Multiple threat and target simulators with high
of-regard sensors update rates are concurrently operated within a
large field of view.
Tailored countermeasures Requires
• a target platform with representative signature
• air and ground threats that appropriately
stimulate the system to determine
countermeasure effectiveness.
Table A.9—continued
Feature Effects on Testing
Integrated maintenance Software intensive, paperless systems require first-

information system and of-kind DT/OT evaluations and assessments.
technical order data
AFMSS and mission support Because these elements of the weapon system have
element a higher level of integration, the testing activity
also must have a higher level of integration.
SOURCE: F-22 TEMP.
tary, civilian, or support contractor), and the remaining 290 were

weapon system contractor personnel. Since the first EMD flight, the
CTF has been collocated at Edwards AFB, California. To identify
operational suitability and effectiveness issues early in testing, DT&E
and OT&E testers will participate in the planning and execution of
military utility testing as part of the combined DT/OT phase of
testing.
Aircraft 4001–4003 are dedicated to performance, structures,
flying qualities, high angle of attack, propulsion, and stores carriage
and separation testing. They are equipped with flight-test nose booms
and provisions for special flight-test instrumentation, such as a flutter
excitation system, a center-of-gravity control system, and a stabiliza-
tion recovery chute. Aircraft 4004–4009 have full avionics suites and
will be used for avionics and weapon integration testing. AFOTEC
will use two EMD aircraft (modified to be production representative)
and two PRTVs for dedicated IOT&E. An additional production-
representative spare aircraft will also be available.
For LO testing, the full-scale pole model was used to finalize the
design and to assess signatures against the RCS specification. In-flight
RCS testing using at least three EMD aircraft will support final signa-
ture verification. Multiple in-flight RCS tests over the test period,
using at least two aircraft, will be used to verify maintenance proce-
dures. AFOTEC test flights will be used both to verify the RCS and
the indications of the RCS status from the signature assessment sys-
tem.
The test program has been rebaselined several times since the
beginning of EMD. To date, approximately 122 test months have
been lost primarily because of late aircraft deliveries for testing. Table
A.10 presents RAND’s projections for the total F/A-22 flight-test
program, based on the flights and flight hours through June 2001
plus the SPO’s estimate of additional flights and flight hours based on
the June 2001 replan.
Table A.11 presents the aircraft calendar months in flight test
(after initial airworthiness testing). Delivery dates for aircraft that had
yet to be delivered are as of January 2002.
Table A.12 reports the DEM/VAL and EMD schedules, along
with sorties, aircraft, and rates, for the F/A-22 planned flight test pro-
gram.
Table A.10
F/A-22 Flight Test; June 2001 Replan
Flight
Test Aircraft Sorties Time
Event (no.) (no.) (hrs)
EMD flight test June 2001– 8a 1,658 3,680

January 2004
Dedicated IOT&E 4 428b 856
April 2003–November 2003
a
Aircraft 4001 retired from flight test before the June 2001 replan.
b
428 scheduled/322 effective sorties, amounting to approximately 28 aircraft months.
Table A.11
The F/A-22 Test Aircraft
Aircraft Calendar
Test Enter Exit Months to Complete
Aircraft EMD Testing EMD Testing a EMD Testing
4001 May 1998 November 2000 31b

4002 August 1998 April 2002 44
4003 September 2000 January 2004 41
4004c January 2001 August 2003 32
4005 March 2001 August 2003 30
4006 May 2001 August 2003 28
Aircraft Calendar
Test Enter Exit Months to Complete
Aircraft EMD Testing EMD Testing a EMD Testing
4007 January 2002 April 2003d 16

4008 March 2002 September 2002 d 7
4009 May 2002 September 2002 d 5
Total 234e
a
SPO projections for completing all EMD testing.
b
Note that 4001 retired before the June 2001 replan.
c
Dedicated to climatic testing for a significant portion of this period. This time has not
been subtracted.
d
These aircraft were assigned to dedicated IOT&E training and dedicated IOT&E, either
as primary or backup, and therefore considered not available for test.
e
This total includes periods when a test aircraft was unavailable for flight test because
of modifications, ground tests, software loading, etc. Some of these times were
planned on June 2001, and some surfaced during execution of the program.
Table A.12
F/A-22 Planned Flight-Test Program
Flight
First Aircraft Sorties Aircraft Time
Flight (no.) (no.) Months (hrs) Ratea
DEM/VAL
October 1986–
January 1991
(YF-22) August 1990 2 74 5b 92a 18.4
September
(YF-23) 1990 2 50 5b 65 13.0
EMDc
August 1991— September
January 2004 1997 9 1,659 234 3,680 10.7
a
In flight hours per aircraft month.
b
Approximate.
c
SPO projections for completing all EMD testing.
T-45 Naval Undergraduate Jet Flight Training System
Mission
The Undergraduate Jet Flight Training System (UJFT) provides
intermediate and advanced strike pilot training using an integrated
ground and flight training system.9 The flight training includes air-
craft familiarization, basic instruments, airway navigation, air-to-
ground and simulated air-to-air weapon delivery, aerial combat
maneuvering, carrier qualifications, low-level navigation, formation
flying, and tactical maneuvering. The system also supports an instruc-
tor training course. The T-45 Training System was developed to
replace both the T-2C and TA-4J.
System Description
The T-45 Training System (T45TS) consists of the T-45 Goshawk
aircraft; an aircraft simulator suite for both instrument and visual
flight training; flight training–related academic materials, including
training courses, equipment, and course materials for UJFT and
instructor training; a computer-based training integration system; and
contractor logistics support.10
The T-45 Goshawk is a tandem-seat single-engine carrier-
capable jet aircraft derived from the existing BAe Hawk. The aircraft
includes the Navy aircrew common ejection seat, the standard atti-
tude heading reference system, an onboard oxygen generating system,
carrier operations capability, and a training weapon delivery capabil-
ity. The T-45 is fully contractor supported, including all levels of
maintenance and logistics.
Programmatics
The T-45 entered EMD in 1984. The initial OT (OT-IIA) in
November 1988 identified major deficiencies in aircraft handling
_____________
9 Information on the T-45 and T-45 Cockpit 21 came from the Test and Evaluation Master Plan No.
786 for the Naval Undergraduate Flight Training System (T45TS) (Revisions 5 and 6), briefing slides
and discussions with the program office.
10 Note that, for this study, we addressed only the aircraft portion of the system.
qualities, which caused several program slips while Boeing (then

McDonnell Douglas) was redesigning the aircraft. OT-IIB in 1990
and OT-IIC in 1991 verified improvement in the deficient areas, and
a successful OPEVAL (OT-IID) was completed in April 1994. The
T45TS was determined to be operationally effective and operationally
suitable.
Test Program
The T-45 was originally planned as a firm-fixed-price demonstration
program that would require only relatively minor modifications. It
was the first modern land-based aircraft to be modified for carrier
capability. However, much of the structure of the Hawk had to be
changed to make it carrier-suitable.
As a result, the program evolved into a major development
effort. Because it used an existing airframe, little M&S was originally
planned; this increased the flight hours necessary to test changes to
the airframe. Because of the performance shortfalls and redesign,
normal aeronautical and engine simulations eventually had to be
developed anyway.
Relatively few government test engineers were assigned to the
program. The Naval Air Warfare Center’s Aircraft Division (NAWC-
AD) did most of the government DT. High-angle-of-attack testing
was done at the Air Force Flight Test Center. Most of the contrac-
tor’s testing was done at its facility in Yuma, Arizona. One test air-
craft was lost in a Class A mishap, resulting in a 12–18 month slip.
Live-fire testing was not required. COMOPTEVFOR conducted the
operational assessments and testing.
All OT periods (and associated flight hours) are government
only, as reported in the T-45 TEMP (see Table A.13). The contrac-
tor flew hundreds of additional flight hours during DT, which the
TEMP does not include. From the beginning of DT to February 1,
1994, which encompasses all DT testing and the first OPEVAL
period, the total contractor and government flight testing was 1,880
flights and 1,932 flight hours. Table A.14 summarizes the informa-
tion in Table A.13.
Table A.13
T-45 Flight Testing
Test Flight
Event (no.) (no.) (hrs) Remarks
DT-IIA 1 17 20.2 Evaluation of flying qualities,

November 6–17, 1988 systems, and performance
OT-IIA 1 10 13.3 Did not support LRIP; program
November 17–21, 1988 restructured
DT-IIA follow-on 1 8 9.8 Further evaluation of flying
February 27– qualities and performance
March 11, 1989
DT-IIA additional follow- 2 3 6.1 Evaluation of human factors,
on June 1990 controls, and displays
DT-IIB November 16– 2 27 27.1 Verification of corrections of prior
December 6, 1990 deficiencies, suitability,
specification conformance
OT-IIB 2 20 24.1 Supported LRIP
December 13–20, 1990
DT-IIC July 23– 2 19 20.0 Evaluation of weapon carriage,
October 7, 1991 high speed flying qualities, ILS,
VOR, lighting, and human
factors
OT-IIC August 1–8, 1991 3 19 22.5 Testing of various training
missions
DT-IID August 13– 2 22 7.0 Land-based catapult and
December 10, 1991 and arrestment
January 3–7, 1992
4 8.2 Initial sea trials; terminated
because of aircraft damage
1 10 8.7 Evaluation of expanded-envelope
flying qualities
DT-IIE 2 25 35.5 Follow-on sea trials
September 1–6, 1993
DT-IIF September 1– 6 47 52.0 High angle of attack testing at
November 23, 1993 Edwards AFB
28 30.7 Navy TECHEVAL
12 27.3 a Follow-on sea trials
OT-IID October 18– 6 583 671.8 System OPEVAL, including
November 17, 1993 and aircraft, academics, Training
February 8–April 8, 1994 Integration System, and
simulators
a
Estimated.
Table A.14
T-45 Testing Summary
Flight
Test Time
Event Aircraft Sorties (hrs) Remarks
DT and first 1–6 1,880 1,932 Contractor and govern-

OPEVAL period ment flight testing;
November 1988–February includes some flights
1994 and hours shown
below under OPEVAL
OT-IID OPEVAL 6 583 671.8 System OPEVAL,
October–November 1993 including aircraft,
and February– academics, training
April 1994 integration system, and
simulators
T-45 Cockpit-21
Mission
The T-45 Training System provides intermediate and advanced strike
pilot training using an integrated ground- and flight-training system.
Undergraduate jet pilot training (UJPT) includes aircraft familiariza-
tion, basic instruments, airway navigation, air-to-ground and simu-
lated air-to-air weapon delivery, aerial combat maneuvering, carrier
qualification, low-level navigation, formation, and tactical maneu-
vering. Also, the system supports an instructor training course. The
T-45TS was developed to replace both the T-2C and TA-4J.
System Description
The T-45 Training System (T45TS) consists of the T-45 aircraft; an
aircraft simulator suite for both instrument and visual flight training;
academic materials, including training courses, equipment, and
course materials for UJPT and instructor training; a computer-based
training integration system; and contractor logistics support. 11 The
_____________
11 Note that, for this study, we addressed only the aircraft portion of the system.
first 83 aircraft delivered to the Navy were T-45As with analog cock-
pits. Subsequent deliveries were T-45Cs, which incorporate a digital
cockpit known as Cockpit-21. To provide an early introduction to
digital displays similar to those used in current fleet tactical aircraft,
the Cockpit-21 avionics upgrade replaces current conventional pri-
mary flight instruments with two multifunctional displays in each
cockpit.
Programmatics
Prototype testing of the Cockpit-21 upgrade began in March 1994,
with the first flight the same month. Approval for fleet introduction
of the T-45C was recommended in December 1998, following OT-
IIIB earlier that year.
Test Program
Contractor and government flight testing ran from March 1994 to
September 1998 and involved up to four aircraft (see Table A.15).
Table A.15
T-45C Flight Testing
Test Flight
Event (no.) (no). (hrs) Remarks
DT-IIG 1 165 215 Combined contractor-

March 1994– government testing of
March 1995 prototype digital cockpit
installation, conducted in
St.Louis and at NAWC-AD
Patuxent River.
System judged not ready for OT;
schedule extended 1 year.
DT-IIIA 1 24 33.2 Combined contractor-
March 6–25, 1996 government testing of
preproduction digital cockpit
installation and verification of
correction of deficiencies (VCD).
Conducted at NAWC-AD
Patuxent River.
System judged ready for OT.
Test Flight
Event (no.) (no). (hrs) Remarks
OT-IIIA 1 28 38.4 OT, including a combined DT/OT

April 9–24, 1996 at sea test period.
Judged potentially operationally
effective and suitable.
DT-IIIB 1 27 31.5 Conducted at NAWC-AD
November 18, 1997– Patuxent River.
February 9, 1998 Recommended that the T-45C
(Cockpit-21) proceed to OT-IIIB.
OT-IIIB 4 602 881.3 The T-45C aircraft was
February 18– determined to be operationally
September 18, 1998 effective and operationally
suitable.
These aircraft flew 846 sorties and flew for 1,199.4 hours. Following
OT-IIIB, the T-45C aircraft was determined to be operationally
effective and operationally suitable.
V-22 Osprey
Mission
The V-22 weapon system is a multiservice, multimission vertical and
short takeoff and landing aircraft.12 The Marine Corps MV-22’s pri-
mary mission is amphibious assault. The Air Force CV-22’s primary
mission is long-range infiltration, exfiltration, and resupply of Special
Operations Forces. Secondary missions are land assault, medical
evacuation, fleet logistics support, and special warfare. The MV-22
will replace the CH-46E and CH-53A/D in the Marine Corps inven-
tory. The CV-22 will replace the MH-53J and MH-60G and will
supplement the MC-130 in the Air Force inventory.
_____________
12 We derived the information in this section from a meeting with the V-22 Program Office July 24,
2001, and from subsequent comments and input; numerous issues of Bell-Boeing’s Tiltrotor Times ; and
the draft V-22 Osprey TEMP No. M960 Rev. B.
System Description
The V-22 is a tilt-rotor aircraft with rotating engine nacelles mounted
on each wingtip, enabling vertical and short takeoff and landing. The
nacelles rotate to the horizontal position for cruising at high speed.
The tilt-rotor design combines the vertical flight capabilities of a heli-
copter with the speed and range of a turboprop airplane and permits
aerial refueling and worldwide self-deployment. Two Rolls Royce
T406-AD-400 turboshaft engines drive two 38-ft. diameter prop-
rotors. The proprotors are connected to each other by an inter-
connecting shaft, which maintains proprotor synchronization and
provides single-engine power to both proprotors in the event of
engine failure. A triply redundant digital fly-by-wire system controls
engines and flight controls. The airframe is primarily graphite epoxy
composite. An integrated EW defensive suite that includes a radar
warning receiver, a missile warning set, and a countermeasures dis-
pensing system will be installed.
The Air Force CV-22 and Marine Corps MV-22 share the same
basic airframe design. The CV-22 is configured for Special
Operations Forces’ infiltration and exfiltration operations. Unique
CV-22 items include a terrain following and terrain avoidance radar
system, additional fuel tanks to increase range, an additional crew
position, an expanded communications suite, and the addition of a
defensive systems suite to enhance survivability during penetration
missions.
Programmatics
V-22 development and testing have taken place in two distinct pro-
grams. An FSD contract was awarded to a Bell-Boeing joint venture
team on May 2, 1986. Six flight-test articles were planned, and five
were completed. First flight of the V-22 in FSD was in March 1989.
The Secretary of Defense cancelled the program on April 19, 1989,
but Congress and the services continued to fund the test program
incrementally until early FY 1993.
Approximately 820 hours of flight testing were completed. Two
aircraft were lost: Aircraft number five crashed on its first flight in
June 1991 because of a problem with its gyroscopic sensors, and air-
craft number four crashed on July 20, 1992, because a section of the
drive shaft failed. On October 22, 1992, the FSD airframe contract
was officially terminated, and a letter contract for EMD was awarded
on the same day.
The purpose of the V-22 EMD program is to design a lighter,
lower cost aircraft than the FSD design and to build four production
articles (aircraft numbers 7–10). An EMD engine contract was
awarded in December 1992. DT and OT continued during EMD
using the three remaining FSD aircraft, supplemented by EMD air-
craft as they were built.
Test Program
During FSD (1986 to 1992), the V-22 T&E program concentrated
on engineering and integration testing performed by the contractor.
NAWC-AD conducted three periods of formal DT, with the partici-
pation of the operational test community in integrated test team
activities. These tests provided early insight into the development
effort.
The EMD flight-test program began by using the FSD V-22
aircraft for design support, risk reduction, and envelope expansion. A
CTF, consisting of Navy, Air Force, Bell, and Boeing personnel, con-
ducted the EMD test program, with the exception of dedicated
operational test events. Testing of four EMD aircraft began in FY
1997, following the first flight of aircraft number 7 in February 1997.
Integrated testing (IT) and OT used a combination of the FSD and
EMD V-22 aircraft during initial assessment (IT-IIA/B/C and OT-
IIA/B/C). OT-IID and subsequent tests used the EMD configuration
aircraft.
The Navy conducted DT&E of the MV-22 at Patuxent Naval
Air Station, Maryland. A CTF stationed at Edwards AFB, California,
conducted the DT&E for the CV-22.
The Secretary of Defense certified a waiver of full system-level
live-fire testing. A comprehensive series of ballistic tests (582 shots
over 16 years) of critical components, major assemblies, and aircraft
structures was conducted. According to DOT&E, live-fire testing was
treated as “an integral part of the design process, not merely as a
method of design verification.” The production design was tested in

444 test firings. Live-fire testing led to a number of design changes.
A multiservice OT team under the direction of COM-
OPTEVFOR conducted the MV-22 and CV-22 OT&E. Initial
CV-22 OT&E culminates with OT-IIH, but six additional MV-22
and CV-22 FOT&E periods have already been identified for
execution to resolve deficiencies from current OT&E, ensure that
deferred OT&E events are finished, and assess P3I elements. CV-22
OT&E consists of several OT periods, designated in accordance with
COMOPTEVFOR conventions as OT-IIA through OT-IIH. For
the CV-22 variant, OT-IIA through OT-IIG are considered opera-
tional assessment periods, with OT-IIH functioning as the equivalent
to IOT&E.
The DOT&E evaluation found the MV-22 operationally effec-
tive but not operationally suitable because of reliability, maintain-
ability, availability, human factors, and interoperability issues. The
CNO had issued a waiver from OPEVAL testing for many V-22
requirements, including combat maneuvering, cargo handling, air-
drop capability, and other capabilities that affect the aircraft’s opera-
tional effectiveness and suitability. DOT&E recommended further
testing of these issues and of the vortex ring state phenomenon, in
which the aircraft loses lift when descending at a low airspeed at too
great a sink rate. One aircraft was lost because it entered a vortex ring
state on an operational training mission in April 2000. When another
fatal mishap occurred in December 2000, flight testing stopped for
17 months. Correction of the vortex ring state problem requires addi-
tional wind-tunnel, digital simulation, and flight testing. An addi-
tional “event-driven” flight-test program began in May 2002.
Tables A.16 and A.17 include data through OPEVAL testing in
July 2000, when the V-22 was judged operationally effective but not
operationally suitable. Testing continued until two fatal mishaps in
2000 caused the grounding of all these aircraft. After lengthy investi-
gations, a new program of flight testing began in May 2002.
Table A.16
V-22 Flight Testing
Flight
Test Sorties Time
Event Aircraft (no.) (hrs) Remarks
FSD Flight Test FSD 653 763.6

March 1989–
October 1992
DT-IIA FSD 9 14.1 Evaluation of flying
March 1990– qualities and
April 1990 performance
characteristics
DT-IIB FSD 15 16.9 Evaluation of flying
November 1990– qualities, performance
December 1990 characteristics, and
shipboard suitability
DT-IIC 2 and 4 20 29.4 Evaluation of readiness for
April 1991– OT-IIA
August 1991 Terminated on loss of
aircraft 5
IT-IIA 2 and 3 92 105.1 Evaluation of envelope
April 1993– expansion
December 1995 Exploration of structural
aero-elastic and flying
qualities
IT-IIB 3 N/A N/A Testing supporting design,
January 1996– risk reduction, and pilot
March 1997 training
IT-IIC 3 and 7 N/A N/A Pilot training and
September 1996– development of
May 1997 preliminary suitability
data
IT-IID EMD N/A 1,469 TECHEVAL to support
December 1996– OPEVAL and determine
September 1998 the final envelope
IT-IIE/F 7 and 91 150 CV-22 testing
February 2000– 10
December 2000
OT-IIA FSD 12 14.8 Testing in support of the
May 1994– Milestone II+ decision
July 1994 and EMD CDR, consisting
primarily of ground tests
and simulations
Flight
Test Sorties Time
Event Aircraft (no.) (hrs) Remarks
OT-IIB FSD 8 10.4 Assessment of operational

June 1995– effectiveness and
October 1995 suitability; primarily
ground tests and
simulations
OT-IIC 3 and 7 N/A 36.1 Assessment of operational
October 1996– effectiveness and
May 1997 suitability and support
LRIP 1 decision; primarily
ground tests and
simulations
OT-IID 9 and 63 142.6 Assessment of operational
September 1998– 10 effectiveness and
October 1998 suitability and support
LRIP 3 decision
OT-IIE, F, G, OPEVAL LRIP a 522 805 Determination of opera-
November 1999– 11–15 tional effectiveness and
July 2000 suitability of the MV-22
Assessment of all CV-22
COIs during three at-sea
periods
Judged operationally
effective but not opera-
tionally suitable
a
28 aircraft months.
Table A.17
V-22 Testing Summary
Aircraft Sorties Aircraft Flight

First Flight (no.) (no.) Months Time (hrs)
FSD March 5 (2 lost) 653 N/A 763.6

May 1986— 1989
October 1992
EMD February 4 and 3 N/A N/A N/A
October 1992— 1997 FSD
May 2000
VCD 8 and 10 2000 to
June 2000 date
(ongoing)
APPENDIX B
Missile Program Descriptions
AIM-9X Sidewinder
Mission
The AIM-9 Sidewinder is a heat seeking air-to-air missile designed for
short-range engagements.1 It complements the medium-range
AMRAAM in the fighter aircraft arsenal. The AIM-9X is a major
modification of the AIM-9, which the U.S. Naval Weapons Center at
China Lake, California, first developed in the 1950s. The missile has
had several variants and is used on all U.S. fighter aircraft, including
the F-15, F-16, and F/A-18, and will be employed on the F/A-18E/F,
F/A-22, and JSF. Many allied nations also use the Sidewinder, and
the Soviets and their allies copied and used its early design.
System Description
The AIM-9M, the AIM-9X’s immediate predecessor, can engage tar-
gets from all aspects; its IR counter-countermeasures and background
discrimination capability are better than those of its own predeces-
sors; and it has a reduced-smoke rocket motor. Deliveries of the M
model began in 1983.
The AIM-9X has improved counter-countermeasures, acqui-
sition range, off-boresight capability, background discrimination,
_____________
1 We derived the information in this section from the AIM-9X Selected Acquisition Report, December
31, 1999; an interview with the AIM-9X Joint Program Office (JPO), May 24, 2001; AIM-9X JPO
(2001a) and (2001b); Sidewinder Missile AIM-9X CARD, Final Update, July 7, 2000, PMA-259; Boe
and Miller (undated).
145
maneuverability, and day or night capability. It uses the same motor,

warhead, and fuze as the AIM-9M and is of similar size and weight.
Component differences include the following:
• A new airframe design and fixed forward wings reduce missile
size and drag. The smaller airframe ensures that the missile will
meet compressed carriage requirements for the F-22 and JSF,
which have internal weapon bays.
• A control actuation system in the rear of the missile provides
thrust vectoring and moveable fins for missile maneuvering. The
tail control is a new development item.
• An improved seeker dome provides improved IR properties.
• A servo-controlled sensor assembly provides high off-boresight
slaving.
• A 128 by 128 focal plane array (instead of the AIM-9M’s single-
element IR detector) produces a digital image for use as a tracker
algorithm and enables the system’s robust IR counter-counter-
measures. This component was developed for Advanced Short-
Range Air-to-Air Missile (ASRAAM).
• An electronics unit provides the AIM-9X with digital signal
processing and tracking capability, IR counter-countermeasure
logic, fly out guidance, and field reprogrammability. The guid-
ance system is a new development item.
A related program, the Joint Helmet Mounted Cueing System
(JHMCS), is being developed concurrently but separately from the
AIM-9X. The JHMCS requires a modified helmet and new aircraft
hardware and software. The JHMCS moves the head-up display to
the helmet and enables slaving aircraft weapons and sensors, such as
the AIM-9X, to head movements. Neither program is required for
the other program to proceed, but AIM-9X and JHMCS are inte-
grated and tested together as much as possible.
Programmatics
The AIM-9X is a joint Navy–Air Force program, with the Navy as
lead service. AIM-9X is a CAIV program, trading off performance
and cost to achieve a best-value solution. A two-year concept-
Missile Program Descriptions 147
exploration phase preceded formal development. In this phase, the

services developed five key performance parameters for the missile:
1. the ability to operate during day or night
2. the ability to operate over land or sea in the presence of IR coun-
termeasures
3. weight, size, and electrical compatibility with all current U.S.
fighters and the F-22
4. the ability to acquire, track, and fire on targets over a wider area
than the AIM-9M can
5. a high probability that a missile launched will reach and kill its
target.
The acquisition strategy involved a competitive two-contractor
DEM/VAL phase, then downselection to one EMD contractor.
Hughes and Raytheon were awarded DEM/VAL contracts in
December 1994 to develop preliminary system designs and to con-
duct seeker demonstrations. DEM/VAL was completed in June 1996.
After the 18-month competitive DEM/VAL program and evaluation
of EMD and LRIP proposals, Hughes (now part of Raytheon) was
selected to be the prime contractor for the AIM-9X missile develop-
ment in December 1996. The contractor has total responsibility for
system performance, including development, production, and life-
time maintenance support. The acquisition philosophy is intended to
motivate the contractor to achieve cost and quality objectives by
including both a 10-year missile warranty and award fees for reducing
missile production costs. The EMD effort was scheduled to end in
2002. A 17-year production period is planned to buy a total of
approximately 10,000 missiles.
The use of several existing components from the AIM-9M,
including the warhead, rocket motor, and fuze, during AIM-9X
development helped reduce technical risks. The AIM-9X design also
includes such nondevelopmental items as the airframe and the engine
control system, which the Air Force had previously developed and
tested. These risk-reduction hardware programs demonstrated that a
tail-controlled missile with small fins would have better performance
than the AIM-9M.
Other technical risks that might have affected EMD were

reduced during the 18 months of DEM/VAL; these included the
seeker and tracker software, new components (such as the sensor and
the guidance and control system), and other critical systems. Because
of the success of the DEM/VAL program, both the contractor and
the program manager considered the AIM-9X missile’s overall techni-
cal risk to be low when it entered EMD. Nevertheless, there were dif-
ficult areas of development, with some technical risk, such as the
development of guidance and control software for use in a counter-
measures environment.
Test Program
Table B.1 summarizes the overall test program (as of June 2002),
which is described in greater detail in the paragraphs below.
Table B.1
AIM-9X Testing
Guided
Event Launches Remarks
DEM/VAL Two competitors
December 1994–
June 1996
DT-IIA 0 More than 50 captive-carry
January 1997– sorties
August 1998 Ground launch of
preprogrammed test vehicle
DT-IIB/C 9 More than 300 captive-carry
September 1998– sorties
August 2001 16 SCTV launches
DT-IID (TECHEVAL) 3 More than 75 captive-carry
March–December 2001 sorties
OT-IIA 5
September 1999–
July 2000
DT Assist 4 Concurrent with TECHEVAL
August–
December 2001
OT-IIB (OPEVAL) 22
August 2002–
May 2003
DT included many different types of test missions and missile

configurations to clear the launch platform flight envelope; to dem-
onstrate the missile system’s performance and flight worthiness; and
to test the integration of missile, launcher, aircraft, and JHMCS. The
missile configurations are described below:
• Instrumented Round with Telemetry (IRT). The AIM-9X
IRT was designed to evaluate the captive-carriage environment
and its effect on the AIM-9X. Approximately 10 missions were
required for each aircraft type. Missile strain gauges, accelerome-
ters, and thermocouples gathered environmental data during the
IRT test missions. The detailed test aircraft matrix included a
broad range of test conditions, environments, and configura-
tions. The IRT missions took place at the beginning of the test
program to allow early modification and improvement of system
reliability before the much-longer captive-carry seeker develop-
ment phase of the test program. The IRT flight-test phase for
the F-15C and F-18C/D was completed by August 1999.
• Separation Control Test Vehicle (SCTV). The first launch of
the AIM-9X SCTV was in March 1999. This vehicle was
designed to verify safe separation of the missile in different
launch conditions from AIM-9X–capable aircraft stations on the
F-15C and F-18C/D. For safety and to verify predicted missile
performance, dynamic pressure and g forces were built up
gradually.
Before a guided launch could take place, a successful SCTV
launch of equal or greater safe-separation difficulty was required.
The F-15Cs and F-18C/Ds made 16 launches to verify the mis-
sile’s envelope.
In addition to flight envelope clearance, SCTV shots provided
guidance and control data to validate models and simulations.
Photo chase aircraft, test aircraft modified to carry high-speed
separation cameras along with ground cameras, captured the
majority of the safe-separation data. Missile fly-out was verified
primarily through missile telemetry and radar and optical
tracking.
• Engineering Development Missile (EDM). The AIM-9X

EDMs were built as either a free-flight version (guided launch)
or a captive unit, the EDM captive-test unit (EDM-CTU).
Numerous captive missions preceded each guided launch.
The EDM-CTU was designed to verify missile seeker tracker
performance, verify software performance, and validate models
and simulations. The EDM-CTU was subjected to multiple
environments, background conditions, lighting conditions,
countermeasures, acquisition ranges, boresight and aspect
angles, and targets. The missions included autonomous missile
acquisition and tracking, radar slaving, and slaving with the
JHMCS in maneuvering and nonmaneuvering environments.
The seeker test matrix was divided between the test centers to
maximize the data gathered. Data collection included aircraft
display videos (head-up display, radar, etc.) and the missile
seeker image. Typically, the seeker image was recorded onboard
the aircraft and telemetered to the control room. In addition to
providing seeker performance data, EDM-CTUs demonstrated
and verified weapon system integration between the AIM-9X,
aircraft, launch rail, and JHMCS. The EDM-CTU also gath-
ered captive-carriage reliability data. The majority of EDM
testing was completed by June 2000.
• Production Representative Missile (PRM). The AIM-9X
PRM, like the EDM, consists of both a free-flight version
(guided launch) and a captive unit, the PRM-CTU. The PRM
has improved hardware for greater image-processing capability
compared to the EDM. Testing and data collection are almost
identical to the EDM process. This phase began in March 2000.
In EMD, DT&E consisted of three dedicated phases, DT-IIA,

DT-IIB/C, and DT-IID. To the extent practical, the testing involved
production-configured missiles built on production tooling. Exten-
sive M&S, captive-carry flights, and live firings verified achievement
of required performance. The three phases were
• DT-IIA (January 1997–August 1998) focused on risk reduction
through collection of IR imagery for seeker development and
development and refinement of M&S. Wind-tunnel testing

supported the six-degrees-of-freedom AIM-9X model. Over 50
captive-carry flights and a ground launch of a preprogrammed
control test vehicle were conducted.
• DT-IIB/C (September 1998–August 2001) included laboratory
testing, over 300 captive-carry flights, 16 SCTV launches, and
nine guided launches. The guided launches focused on demon-
strating missile operation and performance, the aircraft and
launcher interface, and validation of M&S results.
• DT-IID (March 2001–December 2001) served as the AIM-9X
TECHEVAL and included laboratory testing, three guided
PRM launches, and approximately 75 captive-carry sorties.
The OT of the AIM-9X before Milestone III likewise consisted
of three phases:
• OT-IIA (September 1999–July 2000) was an operational

assessment of potential operational effectiveness and suitability
before the LRIP decision. Five EDM launches were conducted
to accomplish mutual DT and OT objectives. In addition, six
PRMs were flown in a captive carry reliability assessment from
August 2000–September 2002.
• DT Assist (August 2001–December 2001) was a second phase
of operational assessment, concurrent with the final months of
TECHEVAL, involving captive carry evaluations and four PRM
launches.
• OT-IIB (August 2002–May 2003) will be the AIM-9X
OPEVAL. Navy and Air Force personnel conducted these
operations, which were scheduled to fire 22 PRMs. COM-
OPTEVFOR and AFOTEC used M&S to evaluate live-fire
performance and performance under conditions when live
operations are not practical.
The AIM-9X paid for initial integration and certification of the
AIM-9X on the F-15C for the Air Force and the F-18C/D for the
Navy as part of the development program. Additional platform
programs—the F-22, F-16, F-18E/F, and other aircraft—will pay for

their own certification.
AIM-120 Advanced Medium-Range Air-to-Air Missile
Mission
AMRAAM is an all-weather, radar-guided, air-to-air missile that
replaced the AIM-7 Sparrow. 2 The F-15, F-16, and F/A-18 currently
carry this missile worldwide, and the F/A-18E/F, F-22, and JSF will
carry it for use against air threats in a variety of weather and elec-
tronic combat conditions. The U.S. Marine Corps Complementary
Low-Altitude Weapon System will use AMRAAM in a surface launch
role. FMS platforms include the German F-4F, the Swedish Gripen
and Viggen, the United Kingdom Tornado and Sea Harrier, and the
Norwegian Advanced Surface-to-Air Missile System (a ground-
launched version of AMRAAM). AMRAAM is capable of intercept-
ing maneuvering and all-aspect targets within and beyond visual
range and allows a single-launch aircraft to engage multiple targets
with multiple missiles simultaneously.
System Description
The AMRAAM weapon system includes the missile, launcher, the
fire-control system, supporting aircraft avionics, and aircrew inter-
faces.
AMRAAM can be either rail or ejector launched and interfaces
electrically with the platforms via the launch stations. The rail
launcher permits firing from wing-mounted stations and is compati-
ble with both AMRAAM and AIM-9. The ejector launcher, which
permits firing from F-15 and F/A-18 fuselage stations, is compatible
with both AMRAAM and AIM-7.
_____________
2 We derived the information in this section from Mayer (1993), information from the AMRAAM JPO
(AAC/YAF), October 17, 2001 through November 18, 2002; Advanced Medium Range Air-to-Air
Missile TEMP, March 2002.
The aircraft normally updates the missile in-flight through a

data link; however, the missile design does not absolutely require this
interface for successful operation at shorter ranges or against non-
maneuvering targets.
AMRAAM system has four guidance modes:
• command-update data link at longer ranges, with active terminal
guidance
• inertial guidance with active terminal if command update is not
available
• active terminal with no reliance on the aircraft’s fire-control sys-
tem at distances within the seeker’s acquisition range
• active radar with home-on-jam during any phase of flight.
The missile’s active radar permits the launch aircraft to engage multi-
ple targets and to use “launch and leave” tactics. The AMRAAM is
propelled by a solid-fuel, reduced-smoke rocket motor.
Programmatics
The AMRAAM acquisition strategy involved a two-contractor
DEM/VAL phase with downselection to a single FSD contractor. A
leader-follower approach was implemented during FSD to promote
competition by the fourth production lot. The AMRAAM began
FSD in December 1981 with a fixed-price-incentive contract award
to Hughes Aircraft Company. In July 1982, Raytheon won the sec-
ond source contract. Milestone II approval was granted in September
1982. At OSD direction, the AMRAAM Producibility Enhancement
Program (APREP) was initiated to lower the production cost of the
missile by identifying and incrementally redesigning high-cost com-
ponents. Congress imposed caps on both FSD and procurement cost.
Milestone IIIA was approved in June 1987 and full-rate production
(Milestone IIIB) in April 1991.
The AIM-120B missile is the result of the APREP. Improve-
ments include a new digital processor, field reprogrammable memory,
and insertion of large-scale integrated circuit and very-large-scale
integrated circuit electronic components. The AIM-120B was intro-
duced late in lot 6.
A P3I program developed the AIM-120C to provide continuous

improvement in missile performance, this program proceeded in
three phases, each with its own hardware and software development
plan:
• P3I Phase 1 developed the AIM-120C-3, using the APREP pro-
gram as a baseline and including compressed-carriage and elec-
tronic counter-countermeasure enhancements. Testing began in
October 1993. Phase 1 flight testing was complete in December
1994. Production was shifted to the AIM-120C-3 in Lots 9–10
(1997–1999).
• P3I Phase 2 began in June 1994 and improved on the electronic
protection and enhanced weapon effectiveness of the phase 1
(AIM-120C-3) missile. Two software tape cut-ins in lots 9 and
11 improved electronic protection. Lot 11 (AIM-120C-4)
included an improved warhead, and the rocket motor for lot 12
(AIM-120C-5) was more than 5inches longer than that of the
phase 1 missile. Lot 13 included a quadrant-sensing target detec-
tion device, yielding the AIM-120C-6.
• The P3I Phase 3 EMD program began in October 1998 with
the objective of further improving guidance and electronic
protection and developing the AIM-120C-7 variant.
In summary, the following are the AMRAAM production lots
and variants produced in them:
• Lots 1 through 5 and early lot 6 produced the AIM-120A. This
basic AMRAAM variant cannot be reprogrammed.
• Lots 6 through 8 produced the AIM-120B, which is repro-
grammable, for the Air Force and Navy. This variant also con-
tinued to be produced for FMS through lot 14 (as of FY 2000).
• Lots 9 and 10 produced AIM-120C-3, which has clipped wings
and fins to allow it to fit inside the F-22 weapon bay and incor-
porates electronic counter-countermeasure enhancements.
• Lot 11 produced the AIM-120C-4, which has an improved war-
head.
• Lot 12 produced the AIM-120C-5, which features an extended

rocket motor. Note that this variant continued to be produced
for FMS customers through lot 15.
• Lots 13 through 15 produced the AIM-120C-6 (FY 1999–2001
funding), which has a new target detection device.
• Lot 16 and subsequent lots will produce the AIM-120C-7,
which will have the same warhead, target detecting device,
rocket motor, and control section as the AIM-120C-6 but will
involve software changes and hardware modifications to the
guidance section.
Test Program
Table B.2 summarizes the overall test program, which we discuss in
more detail in the following paragraphs.
Table B.2
AMRAAM Testing
Live
Event Firings Remarks
DEM/VAL 7 One STV, three CTVs, and 3 GTVs were fired.
November 1979–
December 1981
FSD (DT&E/IOT&E) 100 Ninety AAVIs, 4 AAVs with warheads, and
October 1986–June 1990 six SCTVs were fired.
AF FOT&E (1) 37 The AFOTEC report supported full-rate
February 1990–May 1993 production.
Initial Navy OT-IIIA 6 COMOPTEVFOR concluded that the missile
January–July 1991 was potentially operationally effective and
suitable.
Navy OT-IIIB (OPEVAL) 29 COMOPTEVFOR concluded that the missile
September 1991–May 1994 was partially operationally effective and
suitable and supported IOC on the F/A-18.
AF FOT&E (2) 39 Used lot 4, 5, 6, and 7 missiles to evaluate
June 1993–March 1996 AIM-120A software in electronic attack,
chaff, and multiple-target scenarios.
Testing also included an initial evaluation
of the AIM-120B production software and
a 7,800 hour CCRP. At least one ACE (~39)
was flown for each launch.
P3I Phase 1 DT&E 12 Two SCTVs, one JTV, and nine AAVIs were
October 1993–October fired, and 36 development and preflight
1995 ACE missions were flown.
Table B.2—continued
Live
Event Firings Remarks
AF FOT&E (3A) 26 Verified the operational effectiveness and
August 1996–August 1999 suitability of the AIM-120B/C-3 hardware
and software; verified the correction of
deficiencies; and completed deferred
OT&E. Testing include 26 AAVI flights,
3,712 hours of captive-carry testing for
CCRP, and 25 ACE missions.
P3I Phase 2 DT&E 6 Included Phase 3 risk-reduction testing.
June 1994–November 2001 Testing included 54 ACE flights; digital and
hardware-in-the-loop simulations; and
launches of one SCTV, 13 AAVIs, and one
AAV (with warhead).
FDE (3B) 21 Testing included 16 AAVIs from lots 12 and
September 2000– a 13, five captive-carry reliability vehicles
(reconfigured lot 8 missiles), and a CCRP.
Ten ACE missions supported the launches.
P3I Phase 3 DT&E 13 Testing included eight AAVIs and one
February 2002–August reliability test vehicle, seven ground and
2003 five flight ITV missions, one IMV, eight
prelaunch ACE missions, and 30 ACE
missions to support software
development.
a
Ongoing at reporting time, with an estimated completion date of August 2002.
Demonstration and Validation. During DEM/VAL, Hughes

test-fired 38 instrumented measurement vehicle (IMV) and captive-
load vehicle (CLV) missions from F-14, F-15, and F-16 aircraft and
approximately 46 seeker test unit (STU) missions; one separation test
vehicle (STV); one preshot guided test vehicle (GTV); and, finally,
three GTV productive missions. Raytheon fired 31 IMV/CLV mis-
sions from F-14, F-15, and F-16 aircraft; 15 RAYSCAT missions, 3
two preprogrammed CTV launches, six preshot GTVs, 31 IMV/CLV
missions, and one GTV. At the direction of the SPO, Hughes fired
three more GTVs after DEM/VAL was over, using leftover
DEM/VAL assets.
Full-Scale Development. Combined DT&E/IOT&E took
place throughout the FSD phase, using both FSD and lot 1 produc-
_____________
3 The RAYSCAT was a Raytheon test vehicle that was used to collect seeker waveform, cluster
discrimination, and clutter data.
tion missiles. The combined DT&E/IOT&E began in October 1986

and was complete in June 1990 (the final FSD launch was in January
1989) and consisted of 79 DT&E and 21 IOT&E flights. This
approach was used to achieve an earlier initial operational capability
(IOC). Testing included 90 AMRAAM Air Vehicle Instrumented
(AAVIs) missiles, four AMRAAM Air Vehicles (AAVs) (for warhead
shots), and six Separation Control Test Vehicles (SCTVs).
AMRAAM Producibility Enhancement Program. The
APREP Block I effort was designed to reduce unit production costs
while maintaining baseline performance. The program introduced
alternative subsystem and component designs, vendors, and produc-
tion techniques, with production cut-ins occurring in lots 3, 4, and 5.
APREP Block I flight testing occurred between September 1991 and
March 1992. Program objectives were to demonstrate form, fit, func-
tion, and interface compatibility with the next-higher level of assem-
bly to ensure that introduction of these projects would not degrade
system performance. Tests of AMRAAM captive equipment (ACE)
and simulations were conducted to demonstrate seeker performance.
APREP Block II. APREP Block II flight testing began in Feb-
ruary 1992 and ended in December 1993, with a total of 62 ACE
flights for hardware validation and software development. Nine Block
II–configured missiles (AAVIs) from lot 3 were launched.
FOT&E Phase 1. AFOTEC conducted FOT&E(1) from
February 1990 to May 1993, including evaluating the following:
• fixes incorporated into AIM-120A lot 2 missiles, through a
captive-carry reliability program (CCRP)
• six Desert Storm missiles that had operational captive-carry time
• the effectiveness of lot 2 and 3 missiles in the operational envi-
ronment
• lot 4 software capabilities, including CCRP
• the effectiveness of lot 2 through 4 missiles in multiple target,
electronic attack, chaff, and maneuvering target scenarios.
AFOTEC launched 37 missiles during FOT&E Phase 1.
Initial Navy OT&E (OT-IIIA). The Navy conducted OT-IIIA
from January 1991 to July 1991 to support a Milestone IIIB full-rate
production decision. F/A-18C/D aircraft fired six lot 2 AIM-120A

missiles.
OT-IIIB Navy OPEVAL. To support the missile’s introduction
into its F/A-18C/D fleet, the Navy conducted an OPEVAL of the
AIM-120A from carriers off Point Mugu, California, as well as other
test sites, under varying environmental conditions. The aircraft cap-
tive carried and fired 29 missiles, 17 of which were equipped with
warheads. The Navy also conducted an extensive CCRP, including
carried arrested landings and catapult launches.
FOT&E Phase 2. The U.S. Air Warfare Center (now the 53rd
Wing) conducted FOT&E(2) from June 1993 to March 1996 to fur-
ther test AMRAAM’s operational capabilities. These tests evaluated
the missiles operational effectiveness and suitability in tactically realis-
tic scenarios. Tests included an extensive CCRP and an initial evalua-
tion of the first production software for the AIM-120B. Lot 4, 5, and
6 missiles were used to evaluate improved AIM-120A software in
electronic attack, chaff, and multiple-target scenarios and to identify
operational capabilities and deficiencies.
P3I Phase 1. Testing of P3I Phase 1 (AIM-120C-3) began in
October 1993 and included wind-tunnel testing at AEDC and IMV
flights to quantify the F-15 and F-18 environments, with an F-15
and F-16 each launching one SCTV. ACE (a total of 36 productive
missions) and simulations accommodated software testing. Nine
AAVIs were launched to confirm that the improvements did not
compromise baseline performance and three to test the improvements
themselves.
P3I Phase 2. Phase 2 testing included 54 ACE flights, one
SCTV to test the new control section, and one AAV to test the AIM-
120C-4 configuration with the new warhead for lot 11. Thirteen
AAVIs were launched to test the upgraded control activation system
and the 5-inch rocket motor from lot 12. AAVI launches also tested
software tape 7A, which was cut in to lot 9; 7B, which was cut in to
lot 11; and 13C, which was cut in to lot 13. Four of the 13 AAVIs
tested the new quadrant-sensing target-detection device for lot 13.
FOT&E Phase 3A. The Air Force’s 53d Wing and the Navy’s
Air Test and Evaluation Squadron 9 (VX-9) conducted FOT&E(3A)
from April 1996 to June 2000 using operationally realistic scenarios

at the Eglin Gulf Test Range, Eglin AFB, Florida; the White Sands
Missile Range, New Mexico; the Utah Test and Training Range; and
the NAWC-WD Sea Range, Point Mugu, California.
The testing served to verify the operational effectiveness and
suitability of AIM-120B/C hardware and software updates,4 to verify
that deficiencies had been corrected, and to complete deferred or
incomplete OT&E. The tactical scenario for each launch, coupled
with threat-representative targets and electronic-attack conditions,
ensured operational realism. The team launched a total of 26 missiles:
six for tape 7, revision 6, profiles; three for tape 5, revision 3, profiles;
and 10 for tape 7, revision 7. The Navy launched the remaining
seven missiles for its own profiles. The CCRP used ten of the mis-
siles. ACE missions and preflight simulations preceded each launch to
examine additional missile capabilities (25 productive missions).
Force Development Evaluation (FDE) 3B. FDE(3B) began in
September 2000. The test plan incorporates the following:
• live launches for lot verification
• periodic live launches for AIM-120 software validation and
regression
• captive carry testing, suitability analysis, and live launches of
projected hardware modifications
• ACE missions and computer simulations for further validation
and evaluation of missile capabilities and performance, in
response to inquiries from field users.
The test program consists of 21 launches of missiles from lots 12
and 13; reconfiguration and launching of five (included in the 21
launches) captive-carry reliability vehicles from lot 8, which were
reconfigured and fired for additional validation. COMOPTEVFOR
will conduct a CCRP to establish a baseline for the F/A-18E/F. Ten
ACE runs will support the live launches.
_____________
4 The items of interest were the hardware updates for lots 7–10 hardware; software update tape 7,
revision 6, and tape 7, revision 7, for the AIM-120C; and software update tape 5, revision 3, for the
AIM-120B.
P3I Phase 3. Phase 3 system testing began in February 2002

with ACE flights to support software development and is planned to
be complete in about August 2003. DT&E objectives will support
OT testing. Verification of tactical performance for the AIM-120C-7
test program will rely heavily on captive flight testing (ACE) and
simulation to augment the limited number of missile firings. Ray-
theon will use its SIMFAX hardware-in-the-loop facility to conduct
multiple test runs in a simulated flight environment using actual mis-
sile hardware and software for software development, for validation of
hardware models used in the Tactical AMRAAM Simulation, and for
preflight risk assessments before ACE flights and missile firings.
A planned 38 ACE captive-carry missions will provide data to
support model and simulation validation and to obtain tactical per-
formance verification data. Of these missions, 30 are for software
development and to test the rehosted high-order-language software.
Eight AAVI firings (and one backup) are planned for collecting
data to support model and simulation validation and to obtain end-
to-end missile performance verification in a realistic environment.
Eight preflight ACE missions will occur before each AAVI mission.
Also, an IMV will be used early in the test program to measure vibra-
tion levels to expose deficiencies in chassis design before moving to
the final design. A reliability test vehicle will be used to measure ship-
board shock for deployment of the AIM-120C-7 in fleet operations.
Joint Air-to-Surface Standoff Missile
Mission
JASSM is a precision standoff weapon for attacking heavily defended,
high-priority targets.5 JASSM can be launched beyond the range of
enemy air defenses and can strike fixed or relocatable targets.
_____________
5 We derived the information in this section from the Lockheed Martin JASSM Web site
http://www.jassm.com/; an interview with the JASSM Program Office on May 30, 2001, and
subsequent communications.
System Description
JASSM is 168 inches long. Its major manufactured parts include a
composite shell, fore and aft covers, tail, wings, fuel tanks, engine,
and warhead. The vehicle has an LO design, and each missile is tested
for its radio frequency signature. The tail and wings fold next to the
body before deployment to reduce space. The 1,000-pound class
warhead can penetrate hard targets, works with an impact or delayed
fuze, and is compliant with Navy requirements for insensitive muni-
tions. The missile is guided by a jam-resistant GPS and ring laser gyro
inertial navigation system (INS) with an IR seeker and pattern
matching autonomous target recognition system that provides aim-
point detection, tracking, and strike in the terminal phase. The con-
trol unit uses commercial electronics.
Programmatics
JASSM is the successor to the terminated Triservice Air-to-Surface
Standoff Missile (TASSM) program. Two 24-month program defini-
tion/risk-reduction (PD/RR) cost-plus-fixed-fee contracts totaling
$237.4 million were awarded to McDonnell Douglas and Lockheed
on June 17, 1996. The PD/RR phase was completed in October
1998.
After the open competition in the PD/RR phase, Lockheed was
selected as EMD contractor in April 1998, and a planned 54-month
EMD program began in November 1998. The Defense Acquisition
Board granted approval for LRIP in December 2001.
Test Program
Table B.3 summarizes the tests conducted for this program. The
paragraphs below describe these in greater detail.
JASSM is an acquisition reform program with no government-
directed DT. The contractor is responsible for planning and execut-
ing DT. The program progresses directly from contractor DT&E to
OT&E. DOT&E and AFOTEC were involved during the request
for proposal stage of the program to lay the foundation for data col-
lection during contractor DT&E. Lockheed’s test program (using
Table B.3
JASSM Testing
Guided Launch
Event Launches Platform Remarks
DT-1 0 F-16 Inert round, seeker guides to
January 2001 target
DT-2 1 F-16 Destroyed air defense target
April 2001
DT-3 1 B-52 Destroyed concrete bunker
May 2001
DT-4 1 B-52 Fuse did not arm, did not
July 2001 detonate
DT-5 1 F-16 Arming problem, did not
September 2001 detonate
DT-6 1 F-16 Destroyed hardened bunker
November 2001
DT-7 1 B-52 Destroyed hardened bunker
December 2001
DT-8 1 F-16 Penetrated hardened bunker
April 2002
DT-9 0 F-16 State of the art anti-jam GPS
July 2002 feature
DT-10 0 F-16 Successful navigation in intense
September 2002 jamming environment
OT-1a 1 F-16
April 2002
OT-2a 0 F-16 Two inert rounds
June 2002
OT-3a 1 B-52
August 2002
OT-4a 2 B-52
September 2002
OT-5a 2 B-52
January–March 2003
OT-6a 1 B-52 One live, one inert launch
March–April 2003
a
As of reporting time, these were planned events.
production-representative systems early in the DT program) sup-

ported data collection for combined developmental and operational
test demands and will lead to an independent government IOT&E
upon certification of readiness for OT by the JASSM Program Man-
ager.
Ground tests included wind-tunnel testing of missile carriage

and separation characteristics; signature tests; hardware-in-the-loop
tests to simulate flight; simulation of the missile’s seeker performance
during autonomous navigation and terminal flight; and all-up round
sled test of missile lethality against hard targets.
Flight tests included using jettison and separation test vehicles to
verify safe separation from F-16 and B-52 aircraft; captive-flight func-
tional testing and captive-carry reliability testing to verify flight
readiness of missiles during prelaunch, simulated launch, and
autonomous en route navigation; assessment of carrier suitability for
JASSMs on F/A-18E/F aircraft during catapult launches and arrested
landings; and flight test of all-up rounds to verify end-to-end system
performance.
The basic premises of the DT program are that test articles will
be nearly production representative and that most of the DT test
events should be end-to-end system evaluations (mission planning
through target damage) in operationally realistic employment sce-
narios using live warheads.
The main purpose of the DT program is to verify the JASSM
system performance specifications. The system performance specifica-
tion values are driven by three key performance parameters: expected
minimum number of missiles needed to kill a target, missile opera-
tional range, and carrier operability.
Early in the EMD program, Lockheed conducted a risk-
reduction free-flight test using a prototype PD/RR air vehicle. Before
the end-to-end system test, two controlled flight tests were conducted
at Eglin AFB, Florida, to determine the missile’s aerodynamic charac-
teristics. Lockheed successfully performed its first DT at White Sands
Missile Range, New Mexico, January 19, 2001. It was the first flight
using a seeker to guide to a target. JASSM flew eight powered DT
flight tests through April 2002, including destruction of hardened
targets. JASSM is planned to have ten DT and ten OT missile shots.
The reduced number of shots (compared to other, similar programs)
is due in part to the combined DT and OT test teams.
AFOTEC will be the lead OT agency for the JASSM program,
with OPTEVFOR being the participating OT agency. OPTEVFOR
will conduct such Navy-unique testing as FA-18E/F launch and

shipboard operability during its FOT&E program.
Joint Direct Attack Munition
Mission
JDAM is an accurate, all-weather low-cost guidance kit for current
inventory 2,000- (Mark 84/BLU-109) and 1,000-pound (Mark-
83/BLU-110) bombs.6 JDAM provides highly accurate weapon
delivery in any “flyable” weather. After release, JDAM can receive
updates from GPS satellites to help guide the bomb to the target.
JDAM is a bomb-on-coordinates system that navigates from release
to the planned target coordinates. JDAM provides the user with a
variety of targeting options, such as preplanned targeting using the
Air Force Mission Support System (AFMSS) or the Navy’s Tactical
Automated Mission Planning System (TAMPS), sensor targeting, or
in-flight retargeting using onboard sensors and manual data entry.
JDAM also allows multiple target engagements on a single-pass deliv-
ery.
System Description
The guidance kit includes an INS augmented by GPS updates and a
tail control system. Target coordinates and other guidance data are
passed to the weapon through a MIL-STD-1760 interface from the
delivery aircraft during weapon initialization. Upon release, autono-
mous guidance is initiated using INS data only. After the JDAM GPS
receiver acquires the GPS satellites, precision GPS position and
velocity data are used to refine the INS data. The guidance unit
attaches to the bomb and, through controlled tail fin movements,
directs the bomb to the target. The JDAM is to be integrated on the
B-1B, B-2, B-52H, F-14B/D, F-15E, F-16C/D, F-18C/D, F-18E/F,
_____________
6 We derived the information in this section from the February 2000 TEMP and AFOTEC reports.
F-22, F-117A, and AV-8B aircraft. The B-52 and F/A-18C/D are the
threshold aircraft.
Programmatics
JDAM is a joint Air Force–Navy program; the Air Force is the lead
service. JDAM is an ACAT identification program under OSD over-
sight. Because JDAM was selected as a defense acquisition pilot pro-
gram, many normal procurement requirements were reduced in favor
of using “best commercial practices.” Government management,
oversight, and decision processes were also streamlined.
Four contractors were involved in pre-EMD activities. The
development program had two phases. The Phase 1 EMD effort
began in April 1994 and involved two competing contractors,
McDonnell Douglas and Martin Marietta. The primary focus of
Phase 1 EMD was to reduce manufacturing risks and the projected
average unit production price by having each competitor develop its
design through CDR. The Phase 2 EMD effort began in October
1995 with the selection of one contractor, McDonnell Douglas. The
Phase 2 EMD effort completed system development with emphasis
on development and OT. The Phase 2 EMD contract was a cost-
plus-award-fee contract valued at approximately $102 million with a
period of performance from October 1995 to February 1999. The
Phase 2 EMD contract also included minimum and economic order
quantity options for production lots 1 and 2.
Test Program
Table B.4 summarizes the test program, which we discuss in greater
detail in the following paragraphs.
A test IPT that included representatives from the joint program
office, 46th Test Wing, NAVAIR, and Boeing, the system contractor,
managed DT. Representatives from the OT activities were also
involved early in DT.
Testing in EMD Phase 1 consisted of wind-tunnel testing, digi-
tal simulation, fit checks, use of weapon simulators to test interfaces
between the aircraft and the weapon, supportability analyses, and
Table B.4
JDAM Testing
Units
Airframe Event Fired Remarks
F-18C/D Mark 84 Safe separation 27 STVs Wind-tunnel and safe-
1–4Q FY1996 separation testing
DT&E 48 GTVs Captive carry and release,
1Q FY1997– carrier suitability, aircraft
3Q FY1998 integration, and mission
planning
DT/OT-IIA 14 GTVs Captive carry and release
July 1998– Flew 14 hours total with a
October 1998 restricted flight envelope
because of problems
with the tail actuator
subsystem
Concurrent DT evaluated
tail actuator subsystem
fixes
OT-IIB (OPEVAL) 55 GTVs Independent phase of OT;
November 1998– tested all weapon and
March 1999 fuze combinations
Flew approximately 300
hours (188 sorties)
Verified correction of tail
actuator subsystem
problems
DT/VCD 10 GTVs
1–3Q FY2000
B-52H Mark 84 Safe separation 20 STVs Wind-tunnel and safe-
3Q FY1996 separation testing
DT&E 16 GTVs
2Q FY1997–
2Q FY1998
DT/OT-IIA 16 GTVs
4Q FY1998
OT-IIB IOT&E 40 GTVs Independent phase of OT
1Q FY1999– Tested all weapon and
4Q FY1999 fuze combinations
19 sorties
F-16 Safe separation 25 STVs Wind-tunnel and safe-
3Q FY1996 separation testing
F-16 MK84
DT&E 64 GTVs
2Q FY1997–
2Q FY1998
Units
Airframe Event Fired Remarks
DT/OT-IIA 2 GTVs
4Q FY1998
OT-IIB IOT&E 2 GTVs Independent phase of OT
1Q FY1999– Tested all weapon and
4Q FY1999 fuze combinations
19 sorties
NOTES: The STV is a production-representative airframe with appropriate mass prop-
erties but without a guidance control unit or tail actuator subsystem. The GTV is a
production JDAM guidance kit that may or may not have telemetry, a warhead, or a
fuze.
RCS testing. Flight testing was limited to instrumented measurement

vehicles to define the environment to be expected aboard each type of
aircraft.
The purposes of EMD Phase 2 DT were to demonstrate aircraft
compatibility, safe separation, maneuverability, accuracy, reliability,
maintainability, supportability, and mission planning.
Because the existing general-purpose bombs to which JDAM is
fitted did not change, LFT&E was not necessary.
AFOTEC as the lead OT agency conducted an MOT&E of the
JDAM in conjunction with the Navy’s OPTEVFOR. This MOT&E
consisted of Air Force and Navy combined DT and OT, Air Force
dedicated IOT&E with the B-52, and Navy OT-IIB OPEVAL on
the F-18. All phases were structured to provide operationally realistic
end-to-end mission scenarios, beginning with acceptance inspections
and culminating with inert and live weapon drops. The B-52H and
FA-18C/D delivery profiles were operationally representative,
employing single and multiple weapons against single and multiple
targets.
The Air Force and Navy conducted the combined DT and OT
from July through October 1998 at the Naval Air Warfare Station
China Lake, California, range complex. The Air Force conducted its
dedicated IOT&E at the Utah Test and Training Range from
November 1998 through July 1999. Aircrews from the 49th Test and
Evaluation Squadron, operating from Barksdale AFB, Louisiana;
Andersen AFB, Guam; and Minot AFB, North Dakota, flew Air
Combat Command B-52Hs for these tests. The Navy conducted
OT-IIB at Naval Air Warfare Station China Lake, and on three
aircraft carriers from November 1998 through September 2000.
A total of 207 sorties were flown during the MOT&E phase. Of
these, B-52Hs flew 19 sorties (three captive-carriage and 16 weapon-
release sorties) and FA-18C/Ds flew 188(147 captive-carriage and 41
weapon-release sorties). All missions were planned using either the
AFMSS or the Navy TAMPS. The test team evaluated the effective-
ness of these two systems, including AFMSS and TAMPS core; the
JDAM B-52H aircraft, weapon, and electronics; and FA-18C/D
mission-planning modules.
The test agencies rated JDAM as effective but not suitable
because of a combination of demonstrated weapon system perform-
ance against user requirements and the test team’s judgment of mis-
sion accomplishment. JDAM met the user’s requirements for captive-
carriage reliability for both threshold platforms. However, JDAM did
not meet the user’s requirements for storage reliability, mission reli-
ability, and system reliability. In particular, the guidance kit system
did not meet reliability thresholds, and the thresholds for mean time
to load and mission planning time were not met.
To address unresolved and unsatisfactory issues from IOT&E, a
dedicated FOT&E was planned. System reliability will be tracked
and evaluated through FOT&E and lot acceptance testing.
Joint Standoff Weapon
Mission
JSOW is a family of kinematically efficient 1,000-pound class air-to-
surface glide weapon.7 It has LO and provides multiple kills per pass,
preplanned missions, standoff precision engagement, and launch-and-
_____________
7 We derived this information from the JSOW Selected Acquisition Report of December 31, 1998;
DOT&E Annual Report FY 1996; Program Office (PMA 201) input via email dated April 5, 2002.
leave capability against a wide range of targets, day or night and in all
weather conditions. JSOW is used for interdiction of soft or medium
fixed, relocatable and mobile light and heavy armored targets, massed
mobile armored targets, and antipersonnel and air-to-surface threats.
System Description
The JSOW Baseline (AGM-154A) consists of an airframe, a
guidance-and-control system with INS-GPS capability, and a payload
consisting of 145 BLU-97 submunitions. The JSOW/BLU-108
(AGM-154B) is similar, but the payload consists of 6 BLU-108 sub-
munitions. The JSOW Unitary (AGM-154C) is a Navy-only variant
that adds an autonomous imaging IR seeker to the INS-GPS guid-
ance system and a BAe Broach warhead with penetrator capability.
The all-up round is 160 inches long, has a wingspan of 106 inches
when fully deployed, and weighs 1,065 pounds.
JSOW missions are normally preplanned using the Navy
TAMPS or the AFMSS. JSOW interfaces with the aircraft through a
MIL-STD 1553 data bus. The F-18C/D and E/F, F-16C/D, F-15E,
JSF, B-1B, B-2A, and B-52H can employ this weapon.
Programmatics
The JSOW is a Navy-led, joint Navy–Air Force program. A DEM/
VAL contract was awarded in June 1989. The Navy awarded the
JSOW EMD contract in June 1992 to Texas Instruments Defense
Systems and Electronics (currently Raytheon Missile Systems). The
contract option for LRIP was exercised in February 1997. The Navy
approved full-rate production for the AGM-154A in October 1998.
In October 2002, the Air Force withdrew from the JSOW B
program. The Navy completed development but deferred production.
Test Program
Table B.5 summarizes the test program, which the following para-
graphs describe in more detail. Note that the description addresses
only the JSOW Baseline.
Table B.5
JSOW Testing
Guided
DT-IIA
December 1994–March 1995 2
DT-IIB
March 1995–December 1995 10 At China Lake and Point Mugu
DT-IIC
February 1996–October 1996 10 TECHEVAL and LFT&E
USAF DTE
July 1996–August 1996 2 F-16 integration
OT IIA
May 1996–September 1996 6 Operational assessment
OT-IIB
14 OPEVAL
February 1997–June 1997
Judged operationally effective and
operationally suitable
IOT&E
July 1998 2 OT for F-16
The Naval Air Warfare Center China Lake was the lead test
activity. Other facilities used included Point Mugu, NAWC-AD
(Patuxent River), and the Air Force Air Armament Center at Eglin
AFB.
The Air Force began DT&E flight testing JSOW on the F-16 at
Eglin AFB, Florida, in March 1996. AFOTEC conducted an opera-
tional assessment in December 1996, with 46th Test Wing con-
ducting an update in April 1998 in conjunction with DT&E. The
update focused on targeting and weapon platform performance, using
the results of laboratory M&S, captive-flight, and free-flight testing.
Less-than-desirable progress in integrating the baseline JSOW with
the F-16 hindered Air Force testing. The problem was the control-
section locking pins, a subassembly of the JSOW that Texas Instru-
ments did not manufacture.
The U.S. Navy began OPEVAL testing in February 1997, after
successful DT and initial OT programs. Over the entire test program,
42 of 46 shots were successful.
Sensor Fuzed Weapon
Mission
SFW delivers antiarmor munitions to neutralize a large number of
enemy fighting vehicles in massed formations with a limited number
of sorties.8 It can be launched from a variety of altitudes and weather
conditions and in countermeasure environments. Two related pro-
grams, the WCMD and JSOW, can be used with SFW to give it
greater accuracy and range, respectively.
System Description
The SFW Cluster Bomb Unit (CBU-97/B) consists of three major
parts: the tactical munitions dispenser, the BLU-108 submunition,
and the projectile. The 1,000-pound-class tactical munitions dis-
penser holds ten parachute-stabilized BLU-108 submunitions. Each
submunition holds four armor-penetrating projectiles, each of which
has an IR sensor, for a total of 40 projectiles per SFW.
After the weapon is released, the tactical munitions dispenser
opens and dispenses the ten submunitions. At a preset altitude sensed
by a radar altimeter, a rocket motor fires to spin the submunition and
initiate an ascent. The submunition then releases its four projectiles
over the target area. The projectile’s sensor detects a vehicle’s IR sig-
nature, and an explosively formed penetrator fires at the heat source.
If no target is detected after a specific time, the projectiles fire auto-
matically, causing damage to material and personnel.
SFW is compatible with the A-10, F-15E, F-16, B-1, B-2
(WCMD only), and B-52 (WCMD only).
_____________
8 We derived the information in this section from the Sensor Fuzed Weapon Test and Evaluation
Master Plan, August 1996, Area Attack Systems Program Office; Sensor Fuzed Weapon Test and
Evaluation Master Plan, August 1999, Area Attack Systems Program Office; Information from Area
Attack Systems Program Office (AAC/YH), Eglin AFB; Sensor Fuzed Weapon Selected Acquisition
Report, December 31, 1998; SFW Operational Testing Summary; Director, Defense Operational Test
and Evaluation; Number of launches from FY 1997 DOT&E Annual Report.
Programmatics
SFW began FSD in 1985. The development program was restruc-
tured in June 1989 because of test failures, schedule delays, and
budget changes. The restructured program included a transition to
production. After successful live-fire testing and a successful IOT&E
from September 1990 to December 1991, OSD approved LRIP in
March 1992. Decision authority was delegated to the Air Force in
1994, and the Air Force approved Milestone III for full-rate produc-
tion in June 1996.
Two Producibility Enhancement Program (PEP) hardware
upgrades were initiated for SFW to reduce costs and improve pro-
ducibility through design improvement. The first, PEP-1, involved
electronic and mechanical changes to the projectile, including the use
of an application-specific integrated circuit that placed most of the
electronic components on a single chip. The PEP 1 contract was
awarded in early FY 1994. The PEP-2 contract was awarded in early
FY 1995, and the program was to redesign the sequencer and altime-
ter into one integrated submunition electronics unit. The PEP-2 pro-
gram was cancelled because of technical problems, and the critical
elements were integrated into the P3I program.
In May 1996 the Air Force awarded a P3I contract for SFW.
The P3I program involves three major improvements: (1) improving
performance against countermeasures, (2) altering the warhead design
to improve performance against softer targets without degrading the
current target-set performance, and (3) raising the radar altimeter
height of function to increase area coverage. The current sensor will
be upgraded from a passive IR sensor only to a dual-mode type with
passive IR and an active laser sensor. This upgrade will allow the sen-
sors to discriminate between thermal and physical profiles of targets,
enhance the sensor’s performance against cooler targets, and improve
its effectiveness against countermeasures. The improved warhead con-
sists of a modified copper liner configured to form both a central
penetrator for hard targets and multiple smaller penetrators for soft
targets.
Test Program
Table B.6 summarizes the test program. DT included LFT&E using
actual military vehicles instrumented for the test. Producibility modi-
fications have necessitated several QT&E phases as well. FOT&E of
PEP-1 was completed in 1998. All objectives were met, and testing
results indicated that PEP-1 changes have not degraded the perform-
ance of the SFW.
Standoff Land-Attack Missile–Expanded Response
Mission
SLAM-ER provides standoff all-weather precision strike from carrier-
deployed aircraft against fixed, high-value land targets and, secondar-
ily, against relocatable stationary land targets and ships. 9 The missile
Table B.6
SFW Testing

DT&E 39 38 missions
December 1988– 14 inert and 25 live rounds
March 1992
LFT&E 20 statically aimed submunitions
June 1990– to test for lethality against
September 1990 target
Immobilized all land combat
vehicles
IOT&E I 36 AFOTEC conducted 30 sorties
September 1990– Assessed multiple kills per pass,
January 1992 effectiveness, and reliability
and supportability
IOT&E II 14 AFOTEC conducted 10 sorties
June 1995– Used LRIP weapons
February 1996 Found to be operationally
effective at low altitude
FOT&E of PEP 12 Air Warfare Center conducted
January 1998– three sorties
August 1998
_____________
9 We derived the information in this section from an interview with SLAM-ER Program Office on July
13, 2001 and from the FY 2000 DOT&E report for SLAM-ER.
is intended to fill the gap between long-range cruise missiles and

short-range free-fall munitions.
System Description
The original SLAM was based on the Harpoon antiship missile, to
which it added a GPS-aided INS for midcourse guidance, a Maverick
imaging IR sensor, and a Walleye data link for man-in-the-loop con-
trol. SLAM-ER is a major upgrade with greater range; reduced sus-
ceptibility to countermeasures; greater capability against hardened
targets; an improved guidance navigation unit; and improved user
interfaces for mission planning, launch, and control. The primary
changes from the existing SLAM included the following:
1. a modified Tomahawk Block III warhead
2. an improved data link with greater range and jam resistance
3. an improved guidance set with integrated GPS-INS, a 1760 data
bus interface, a multichannel GPS receiver
4. modified Tomahawk wings
5. an automated mission-planning system.
Programmatics
SLAM-ER continued the SLAM development approach of maximiz-
ing use of existing components. After cancellation of the TASSM
program, Boeing received an EMD contract in March 1995 to
improve identified deficiencies of the interim SLAM. The first flight
was in March 1997. LRIP I was approved April 1997, LRIP II in
April 1998, and LRIP III in August 1999. Full-rate production was
approved in May 2000, with existing SLAMs to be upgraded to
SLAM-ER configuration. The threshold platform was the F/A-
18C/D.
Test Program
Table B.7 summarizes the test program. There were five DT and
eight OT launches. LFT&E was required because of the use of a new
titanium-cased warhead. LFT&E consisted of confined volume test-
Table B.7
SLAM-ER Testing
Guided
DT-1 1 First flight
March 1997 Verified basic performance, mission planning,
and maneuvering
DT-2 1 Verified terrain following, data link, pilot
October 1997 designation, and target impact
DT-3 1 Verified range for low-level launch and flyout,
December 1997 shift from IR track to designated aimpoint,
and target impact
DT-4 1 Verified range for high-altitude launch, off-
February 1998 axis launch, steep impact angle, and stop-
motion aimpoint update
DT-5 1 Verified new operational flight program,
March 1998 quick-reaction launch and midcourse
update, target ID, and autonomous lock on
a moving ship target
DT/OT-1 1 Verified land-based target-of-opportunity
June 1998 capability
DT/OT-2 1 Demonstrated the ability to attack a hardened
June 1998 aircraft shelter
DT/OT-3 1 Demonstrated the ability to attack a high-
June 1998 altitude land-based target
DT/OT-4 1
September 1998 Consisted of production verification test
OT-IIA (OPEVAL) 6 Six missiles fired in operationally realistic
December 1998– scenarios
May 1999 One prior DT/OT launch included in OT
analysis
Missile judged not suitable and not effective
VCD 1 Verified software and hardware (missile and
June 1999 AN/AWW-13 data link pod) changes
DT-II 1
October 1999
OT-IIB 4 Four missiles fired in operationally realistic
November 1999– scenarios
January 2000 Deficiencies corrected
Missile judged suitable and effective
ATA OT 1 Evaluate ATA capability
February 2002
ing, three arena tests of warhead fragmentation, and four sled tests of
warhead penetration.
Wind-Corrected Munition Dispenser
Mission
WCMD is an all-weather guidance kit that replaces the tail on the
SFW (CBU-97), Combined Effects Munitions (CBU-87), and the
Gator Mine System (CBU-89).10 With the WCMD tail kit attached,
these weapon designations become CBU-105, CBU-103, and CBU-
104, respectively. WCMD is not a precision capability but does
improve accuracy, depending on the quality of the navigation unit
used. It can be launched from a variety of altitudes, weather condi-
tions, and in countermeasure environments. WCMD corrects for
wind effects, ballistic dispersion, and launch transients from moderate
to high altitudes.
System Description
The WCMD kit integrates a low-cost INS, control unit, and
steerable fins to guide tactical munition dispenser weapons. It is a
bomb-on-coordinates weapon and is used by aircraft that have GPS-
quality heading, velocity, and position data. The data are passed from
the aircraft to the weapon via an MIL-STD-1760 interface. The
threshold aircraft for compatibility are the F-16 and B-52. The objec-
tive aircraft are the B-1, F-15E, A-10, F-117, and JSF.
Programmatics
WCMD is an Air Force ACAT II program using commercial prac-
tices and government streamlining as much as possible. The con-
tracting process minimized the use of military specifications. The
acquisition emphasized affordability. Contractors received a perform-
ance requirement and price requirement for the average unit pro-
curement and could trade-off performance objectives to minimize
_____________
10 We derived the information in this section from the Wind Corrected Munitions Dispenser, Test and
Evaluation Master Plan, Milestone III Revision, January 2001; the Wind Corrected Munitions
Dispenser (WCMD) Engineering and Manufacturing Development (EMD) Cost Estimate, January
2001; and discussions with the Area Attack Systems Program Office (AAC/YH).
costs. The resulting average unit procurement cost for 40,000 units
was approximately $10,000 in FY 1994 dollars.
The development program had two phases. EMD Phase 1 con-
tracts for development with an option for pilot production were
awarded in January 1995 to Alliant Techsystems and Lockheed Mar-
tin. The contractors built production-representative tail kits, and the
Air Force held a fly-off competition. In January 1997, the Air Force
selected Lockheed to continue into Phase 2, pilot production. In fall
1997, flight testing revealed an actuator problem during high-speed
releases. The program was restructured to extend EMD with a third
LRIP lot to allow time to design and qualify a fin-locking mecha-
nism. In April 2001, WCMD was approved for full-rate production.
Test Program
Table B.8 summarizes the test program, which we describe in more
detail below.
Table B.8
WCMD Testing
Number
Guided of
Event Launches Flights Remarks
Contractor pre–fly-off tests 11 10 Light testing verified
January 1995– predictions of contractors’
October 1996 six-degrees-of-freedom
models
Fly-off 14 8
November 1996–
December 1996
DT/OT 61 45 Approximately 60 captive-
carry flights
(approximately 120–180
flight hours)
Phase 1
May 1998–
October 1998
Number
Guided of
Event Launches Flights Remarks
Phase 2
October 1999–
December 2000
IOT&E 21 6
January–
October 2000
The contractor used a digital system model capable of simulat-

ing INS characteristics, vehicle aerodynamics, actuators, inertial
measurement units, and weapon operational flight program functions
in various flight conditions, including launch and terminal maneu-
ver.11 Weapon simulator units were used to support lab testing to
integrate the weapon to host aircraft.
Alliant and Lockheed conducted flight tests before the fly-off,
using eight and nine test units, respectively. During the fly-off, F-16s
dropped seven weapons from each contractor at Eglin AFB under
various conditions.
The SPO and AFOTEC worked together to structure a com-
bined DT/OT program for EMD after the fly-off. The test program
had two phases. Phase 1 DT/OT involved the F-16 and B-52 and
low dynamic pressure conditions; dedicated B-52 IOT&E followed.
Phase 2 DT/OT and dedicated IOT&E involved the B-52 and F-16
and used pin-puller-configured hardware. The aircraft dropped
examples of all three weapon types from various altitudes. The test
program verified compliance with all the ORD key performance
parameters, and no deficiencies are unresolved. The 46th Test Wing
at Eglin AFB conducted the F-16 testing, and the 419th Flight Test
Squadron at Edwards AFB conducted the B-52 testing. LFT&E was
not necessary because WCMD does not modify the explosive charac-
teristics of the munitions.
_____________
11 The program office estimates that simulation eliminated at least 12 flights and drops.
APPENDIX C
System Test and Evaluation Work Breakdown

Structure
This appendix presents an extract from the January 2, 1998, edition

of MIL-HDBK-881, System Test and Evaluation.
H.3.3 System Test and Evaluation
The use of prototype, production, or specifically fabricated
hardware/software to obtain or validate engineering data on the
performance of the system during the development phase (nor-
mally funded from RDT&E) of the program.
Includes:
• detailed planning, conduct, support, data reduction and
reports (excluding the Contract Data Requirements List
data) from such testing, and all hardware/software items
which are consumed or planned to be consumed in the con-
duct of such testing
• all effort associated with the design and production of mod-
els, specimens, fixtures, and instrumentation in support of
the system level test program
NOTE: Test articles which are complete units (i.e., functionally

configured as required by specifications) are excluded from this
work breakdown structure element.
179
Excludes:
• all formal and informal testing up through the subsystem
level which can be associated with the hardware/software
element
• acceptance testing
NOTE: These excluded efforts are to be included with the

appropriate hardware or software elements.
H.3.3.1 Development Test and Evaluation

This effort is planned, conducted, and monitored by the devel-
oping agency of the DoD component. It includes test and
evaluation conducted to:
• demonstrate that the engineering design and development
process is complete.
• demonstrate that the design risks have been minimized.

• demonstrate that the system will meet specifications.
• estimate the system’s military utility when introduced.
• determine whether the engineering design is supportable
(practical, maintainable, safe, etc.) for operational use. pro-
vide test data with which to examine and evaluate trade-offs
against specification requirements, life cycle cost, and
schedule.
• perform the logistics testing efforts to evaluate the achieve-
ment of supportability goals, the adequacy of the support
package for the system (e.g., deliverable maintenance tools,
test equipment, technical publications, maintenance instruc-
tions, and personnel skills and training requirements, etc.).
Includes, for example:

• all contractor in-house effort
System Test and Evaluation Work Breakdown Structure 181
• (all programs, where applicable) models, tests and associated

simulations such as wind tunnel, static, drop, and fatigue;
integration ground tests; test bed aircraft and associated
support; qualification test and evaluation, development
flight test, test instrumentation, environmental tests, ballis-
tics, radiological, range and accuracy demonstrations, test
facility operations, test equipment (including its support
equipment), chase and calibrated pacer aircraft and support
thereto, and logistics testing
• (for aircraft) avionics integration test composed of the fol-
lowing:
– test bench/laboratory, including design, acquisition, and
installation of basic computers and test equipments
which will provide an ability to simulate in the laboratory
the operational environment of the avionics sys-
tem/subsystem
– air vehicle equipment, consisting of the avionics and/or
other air vehicle subsystem modules which are required
by the bench/lab or flying test bed in order to provide a
compatible airframe avionics system/subsystem for
evaluation purposes
– flying test bed, including requirements analysis, design of
modifications, lease or purchase of test bed aircraft,
modification of aircraft, installation of avionics equip-
ment and instrumentation, and checkout of an existing
aircraft used essentially as a flying avionics laboratory
– avionics test program, consisting of the effort required to
develop test plans/procedures, conduct tests, and analyze
hardware and software test results to verify the avionics
equipments’ operational capability and compatibility as
an integrated air vehicle subsystem
– software, referring to the effort required to design, code,
de-bug, and document software programs necessary to
direct the avionics integration test
– (for engines) engine military qualification tests and
engine preliminary flight rating tests
– (for ships) model basin, hydrostatic, fatigue, shock, spe-
cial sea tests and trials, etc., including the Extended Ship
Work Breakdown Structure (ESWBS), Trials Agenda

Preparation, Data Collection & Analysis (842); Dock
and Sea Trials (9823); and Hull Vibration Survey (9825)
elements
H.3.3.2 Operational Test and Evaluation

The test and evaluation conducted by agencies other than the
developing command to assess the prospective system’s military
utility, operational effectiveness, operational suitability, logistics
supportability (including compatibility, inter-operability, reli-
ability, maintainability, logistic requirements, etc.), cost of own-
ership, and need for any modifications.
• Initial operational test and evaluation conducted during the
development of a weapon system
• such tests as system demonstration, flight tests, sea trials,
mobility demonstrations, on-orbit tests, spin demonstration,
stability tests, qualification operational test and evaluation,
etc., and support thereto, required to prove the operational
capability of the deliverable system
• contractor support (e.g., technical assistance, maintenance,
labor, material, etc.) consumed during this phase of testing
• logistics testing efforts to evaluate the achievement of sup-
portability goals and the adequacy of the support for the sys-
tem (e.g., deliverable maintenance tools, test equipment,
technical publications, maintenance instructions, personnel
skills and training requirements, and software support
facility/environment elements)
H.3.3.3 Mock-ups
The design engineering and production of system or subsystem
mock-ups which have special contractual or engineering signifi-
cance, or which are not required solely for the conduct of one of
the above elements of testing.
System Test and Evaluation Work Breakdown Structure 183
H.3.3.4 Test and Evaluation Support

The support elements necessary to operate and maintain, during
test and evaluation, systems and subsystems which are not
consumed during the testing phase and are not allocated to a
specific phase of testing.
• repairable spares, repair of reparables, repair parts, ware-
housing and distribution of spares and repair parts, test and
support equipment, test bed vehicles, drones, surveillance
aircraft, tracking vessels, contractor technical support, etc.
Excludes:
• operational and maintenance personnel, consumables, spe-
cial fixtures, special instrumentation, etc., which are utilized
and/or consumed in a single element of testing and which
should be included under that element of testing
H.3.3.5 Test Facilities

The special test facilities required for performance of the various
developmental tests necessary to prove the design and reliability
of the system or subsystem.
• test tank test fixtures, propulsion test fixtures, white rooms,
test chambers, etc.
Excludes:
• brick and mortar-type facilities identified as industrial facili-
ties
Bibliography
Bell, Don G., Naval Weapons Center Test and Evaluation Model (for Air-
Launched Weapons), China Lake, Calif.: Naval Weapons Center,
October 1988.
The Boeing Company, “Boeing 777 Facts.” Online at http://www.boeing.
com/commercial/777family/pf/pf_facts.html (as of January 28, 2004).
Cook, Cynthia, and John C. Graser, Military Airframe Acquisition Costs:
The Effects of Lean Manufacturing, Santa Monica, Calif.: RAND
Corporation, MR-1325-AF, 2001.
Commander Operational Test and Evaluation Force, Operational Test
Director’s Guide, COMOPTEVFORINST 3960.1H, December 13,
1995.
Defense Science Board, Report of the Defense Science Board Task Force on
Test and Evaluation, Washington, D.C.: Office of the Under Secretary of
Defense for Acquisition and Technology, September 1999.
Defense Systems Management College, Test and Evaluation Management
Guide, 4th edition, Fort Belvoir, Va.: The Defense Acquisition Univer-
sity Press, November 2001.
Director, Operational Test and Evaluation, Annual Report to Congress for
Fiscal Year 2000.
DoD—see U.S. Department of Defense.
F-22 System Program Office, F-22 Test & Evaluation Master Plan, Wright-
Patterson AFB, Ohio, June 1999.
185
Gogerty, David C., Bruce M. Miller, J. Richard Nelson, Paul R. Palmer,

Jr., Acquisition of Contemporary Tactical Munitions, Vol. I: Summary
Report, Alexandria, Va.: Institute for Defense Analyses, March 1990.
Jonson, Nick, “Merits of Streamlining DoD Test and Evaluation
Operations Debated,” Aerospace Daily, May 22, 2002.
Lorell, Mark, and John C. Graser, An Overview of Acquisition Reform Cost
Savings Estimates, Santa Monica, Calif.: RAND Corporation, MR-1329-
AF, 2001.
U.S. Air Force, Developmental Test and Evaluation, Air Force Instruction
99-101, November 1, 1996.
_____, Operational Test and Evaluation, Air Force Instruction 99-102, July
1, 1998.
_____, Armament/Munitions Test Process Direction and Methodology for
Testing, Air Force Manual 99-104, August 1, 1995.
_____, Live Fire Test and Evaluation, Air Force Instruction 99-105, July 25,
1994.
_____, Test Resource Planning, Air Force Instruction 99-109, July 21, 1994.
_____, Test and Evaluation: Airframe-Propulsion-Avionics Test and
Evaluation Process, Air Force Manual 99-110, July 3, 1995.
U.S. Air Force Scientific Advisory Board, Report on Review of the Air Force
Test and Evaluation Infrastructure, SAB-TR-97-03, November 1998.
U.S. Department of Defense, Weight and Balance Data Reporting Forms
for Aircraft (Including Rotorcraft), MIL-STD-1374, September 13,
1977.
_____, Department of Defense Handbook: Work Breakdown Structure,
Appendix H, MIL-HDBK-881, January 2, 1998.
_____, Major Range and Test Facility Base (MRTFB), DoD Directive
3200.11, January 26, 1998.
_____, Mandatory Procedures for Major Defense Acquisition Programs
(MDAPS) and Major Automated Information System (MAIS) Acquisition
Programs, DoD 5000.2-R, April 5, 2002.
U.S. General Accounting Office, Best Practices: A More Constructive Test
Approach Is Key to Better Weapon System Outcomes, GAO/NSIAD-00-
199, July 2000.
Bibliography 187
Younossi, Obaid, Mark V. Arena, Richard M. Moore, Mark Lorell, Joanna

Mason, and John C. Graser, Military Jet Engine Acquisition: Technology
Basics and Cost-Estimating Methodology, Santa Monica, Calif.: RAND
Corporation, MR-1596-AF, 2003.
Younossi, Obaid, Michael Kennedy, and John C. Graser, Military Airframe
Costs: The Effects of Advanced Materials and Manufacturing Processes,
Santa Monica, Calif.: RAND Corporation, MR-1370-AF, 2001.

Test and Evaluation Trends and Costs For Aircraft and Guided Weapons

Uploaded by

Copyright:

Available Formats

Test and Evaluation Trends and Costs For Aircraft and Guided Weapons

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Test and Evaluation Trends and Costs For Aircraft and Guided Weapons

Uploaded by

Copyright:

Available Formats

CHILD POLICY This PDF document was made available

CIVIL JUSTICE from www.rand.org as a public service of

HEALTH AND HEALTH CARE Jump down to document6

NATIONAL SECURITY The RAND Corporation is a nonproﬁt

SCIENCE AND TECHNOLOGY

For More Information

Limited Electronic Distribution Rights

Bernard Fox, Michael Boito, John C.Graser, Obaid Younossi

Prepared for the United States Air Force

Library of Congress Cataloging-in-Publication Data

Cover photo courtesy of the U.S. Air Force at www.af.mil.

The RAND Corporation is a nonprofit research organization providing

© Copyright 2004 RAND Corporation

All rights reserved. No part of this book may be reproduced in any

Published 2004 by the RAND Corporation

This is one of a series of reports from a RAND Project AIR FORCE

This monograph should be of interest to government cost ana-

RAND Project AIR FORCE

Force with independent analyses of policy alternatives affecting the

Determine Test Objectives ............................................................................................ 22

Operational Test and Evaluation ................................................................................ 60

Bibliography ............................................................................................................................. 185

2.1. T&E Phases Within the Acquisition Process............................................ 14

1.1. T&E Perspectives.......................................................................................................6

5.16. Summary Characteristics of Missiles and Guided

T&E is a key step in the development of any military weapon system.

To interpret the results of this study correctly, certain limita-

Cost-Estimating Data, Methodologies, and Trends

tors still rely on government test facilities and functional expertise,

Overall T&E Cost Trends

T&E Issues and Findings

nondevelopmental aircraft for DoD applications, was not generally as

Modeling and Simulation

Testing of Software-Intensive Systems

Combined and Multiservice Testing

Contractor Versus Government Test Facilities

limited, we could draw no general conclusions about the real cost of

In performing this study, the authors had extensive discussions with

• Naval Air Warfare Center—Aircraft Division: J. R. Smullen;

Wascavage, Eric Wilson, and Vicky Yoshida, whose insights and

AAC Air Armament Center

CAIV cost as an independent variable

FAA Federal Aviation Administration

SAF/AQ Assistant Secretary of the Air Force (Acqui-

Background: The Cost of Testing

Test and evaluation (T&E) is a key step in the development of any

NOTE: * Program in progress

Changes in the Test Process

achievement of critical performance parameters and minimize the

Role T&E Objective

Managers, who are concerned with schedules and budgets, tend

Estimating the Cost of T&E

strong to better represent the characteristics, and presumably the cost,

advanced concept technology demonstration (ACTD) programs,

Organization of the Report

The T&E Process

T&E involves the collection and analysis of data on the actual or

Intermittent activities of varying scope

Phase Concept & System Production Support