Comanche Helicopter: Testing Needs to Be Completed Prior to Production
Decisions (Letter Report, 05/18/95, GAO/NSIAD-95-112).

GAO reviewed the Army's Comanche helicopter program, focusing on the
cost and technical issues associated with the restructured program.

GAO found that: (1) the past risks associated with the Comanche's
development and production will continue under the Army's restructured
program; (2) production decisions will be made before operational
testing of the Comanche begins and the development phase will be
extended to fiscal years 2002 and beyond; (3) six additional aircraft
will be acquired, which will provide the Army with the opportunity to
conduct operational testing before committing funds to any production
decisions; (4) the Comanche will be much more expensive than planned, as
its unit cost has tripled in the last 10 years due to program
restructuring and a 74-percent decrease in procurement quantities; (5)
the Comanche may not meet its wartime availability requirements and
lower its operating costs, since the Army has been experiencing
technical problems; and (6) the Comanche program is currently meeting
its goals of reducing maintenance levels and keeping the Comanche's
overall weight growth within acceptable limits.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  NSIAD-95-112
     TITLE:  Comanche Helicopter: Testing Needs to Be Completed Prior to 
             Production Decisions
      DATE:  05/18/95
   SUBJECT:  Helicopters
             Military aircraft
             Combat readiness
             Advanced weapons systems
             Military cost control
             Army procurement
             Aircraft maintenance
             Military operations
             Testing
             Aircraft research
IDENTIFIER:  Comanche Helicopter
             Army Comanche Helicopter Program
             T-45A Aircraft
             C-17 Aircraft
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Committees

May 1995

COMANCHE HELICOPTER - TESTING
NEEDS TO BE COMPLETED PRIOR TO
PRODUCTION DECISIONS

GAO/NSIAD-95-112

Comanche Helicopter


Abbreviations
=============================================================== ABBREV

  DOD - Department of Defense
  MTTR - mean time to repair
  MMH/FH - maintenance man-hours per flight hour

Letter
=============================================================== LETTER


B-259389

May 18, 1995

The Honorable Strom Thurmond
Chairman
The Honorable Sam Nunn
Ranking Minority Member
Committee on Armed Services
United States Senate

The Honorable Floyd Spence
Chairman
The Honorable Ronald Dellums
Ranking Minority Member
Committee on National Security
House of Representatives

This report presents the results of our review of the Department of
the Army's Comanche helicopter program.  Because this program is the
centerpiece of Army aviation modernization efforts and faces major
funding shortages and restructuring, we focused our review on cost
and technical issues.  Because of your expressed interest in the
Comanche program, we believe the information in this report will be
useful as you review the Department of Defense's (DOD) plans for the
program. 


   BACKGROUND
------------------------------------------------------------ Letter :1

The Comanche program was established in 1983 to replace the Army's
light helicopter fleet.  The contractor team of Sikorsky Aircraft
Corporation and Boeing Helicopter Company were expected to design a
low-cost, lightweight, advanced technology helicopter capable of
performing the primary missions of armed reconnaissance and attack. 
Critical to achieving these capabilities are the successful
development of advanced technologies, including composite materials,
advanced avionics and propulsion systems, and sophisticated software
and hardware.  The Army must meet ambitious maintainability goals in
order to (1) realize significantly lower operating and support costs
predicted for this program and (2) achieve a wartime operational
availability for the Comanche of
6 hours per day. 

In December 1994, the Secretary of Defense directed the Army to
restructure the Comanche helicopter program as part of efforts to
meet budgetary constraints.  The Secretary's restructure decision
reduced funding for the program from $4.2 billion to $2.2 billion for
fiscal years 1996 through 2001.  In addition to extending the
development phase by
3 years, it also called for two flyable prototypes to be produced and
the Comanche production decision to be deferred. 

In response to the Secretary's decision, the Army proposed a program
restructure that would allow it to acquire, within the Secretary's
funding constraint, six aircraft in addition to the two prototypes by
deferring developmental efforts to fiscal year 2002 and beyond.  DOD
approved the proposal in March 1995. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :2

Under the Army's restructured program, production decisions will be
made before operational testing of the Comanche starts, thereby
continuing the high degree of risks associated with concurrent
development and production.  However, the time provided by extending
the development phase and the acquisition of the six additional
aircraft under the restructure provides the Army with the opportunity
to conduct operational testing before committing funds to any
production decisions.  Because of the Comanche program's high costs
and technical risks, we believe the Army should complete operational
testing before long-lead and low-rate initial production decisions
are made, thereby significantly reducing the risks from the program. 

The Comanche will be a much more expensive armed reconnaissance
helicopter than the one originally justified to the Congress.  The
Comanche's program acquisition unit cost\1 has almost tripled in 10
years.  It has increased from $12.1 million in 1985 to $34.4 million
as of February 1995.  Program acquisition unit cost increases
occurred primarily because of program restructuring and a 74-percent
decrease in the quantity of aircraft to be procured.  Cost and
program schedule will again be affected because of the decisions to
restructure the program. 

After a decade of developing the Comanche, the Army continues to
experience technical problems.  The Army is experiencing software
development and testing problems associated with electronic systems
that affect the performance of the Comanche.  All key aircraft
maintainability requirements for the Comanche may not be achievable. 
Therefore, the Comanche's ability to meet its wartime availability
requirements and its objective of lower operating and support cost is
questionable.  On the positive side, the program is currently meeting
its goals of reducing maintenance levels and keeping within
acceptable limits of overall weight growth for the Comanche. 


--------------------
\1 Throughout this report, both the total program acquisition costs
and program acquisition unit costs are depicted in current dollars
and include research, development, and acquisition costs but exclude
military construction costs. 


   RESTRUCTURING CONTINUES RISKS
------------------------------------------------------------ Letter :3

The Army's restructuring of the Comanche program continues risks (1)
associated with making production decisions before knowing whether
the aircraft will be able to perform as required and (2) of higher
program costs. 


      RESTRUCTURING CONTINUES
      HIGHLY CONCURRENT PROGRAM
---------------------------------------------------------- Letter :3.1

According to DOD's April 1990 guidelines for determining degrees of
concurrency, a program with high concurrency typically proceeds into
low-rate initial production before significant initial operational
test and evaluation is completed.  Regarding the need to keep
concurrency low, the guidelines note that establishing programs with
no concurrency, or a low degree of concurrency, avoids the risks that
(1) production items have to be retrofitted to make them work
properly and (2) system design will not be thoroughly tested.  As we
recently reported, aircraft systems, including the T-45A and C-17,
that entered low-rate initial production before successfully
completing initial operational testing and evaluation experienced
significant and sometimes costly modifications to achieve
satisfactory performance.\2 Under the Army's restructured program,
operational testing will not begin until after the low-rate initial
production decision is made, continuing the risks associated with the
highly concurrent Comanche program. 

In responding to the Secretary's restructure decision, the Army
proposed, and was subsequently granted approval, to buy six "early
operational capability" aircraft, in addition to the two prototypes
that were to be acquired under the Secretary's decision.  According
to program officials, these aircraft are estimated to cost in excess
of $300 million.  The Army does not consider these aircraft as either
prototype or low-rate initial production aircraft; however, program
officials believe that when these aircraft are fielded, the Army will
be able to better evaluate the Comanche's mission capability.  The
Army intends to fund these aircraft by deferring additional
developmental efforts to fiscal years 2002 and beyond. 

Under the Army's restructured program, operational testing will not
begin until well after funds are committed to buy production
aircraft.  Armed reconnaissance and attack mission equipment packages
are to be integrated into the six early operational aircraft by
fiscal year 2004.  The Army plans to use these aircraft to start
operational testing by about August 2005.  However, long-lead
production decisions are scheduled for November 2003, and low-rate
initial production is planned to start in November 2004, about 9
months before operational testing begins. 

According to DOD's guidelines, the amount of risk associated with
concurrency can be limited by reducing production aircraft to the
minimum necessary to perform initial operational testing.  The Army
maintains that under the stretched out program it can conduct initial
operational testing with the six early operational aircraft.  Because
the restructure has provided the additional time and aircraft, the
Army has an opportunity to significantly reduce or eliminate program
concurrency and its associated risks by completing operational
testing before committing funds to any production decisions. 


--------------------
\2 Weapons Acquisition:  Low-Rate Initial Production Used to Buy
Weapon Systems Prematurely (GAO/NSIAD-95-18, Nov.  1994). 


      COMANCHE COSTS CONTINUE TO
      INCREASE
---------------------------------------------------------- Letter :3.2

The Comanche was originally justified to the Congress as a relatively
inexpensive aircraft.  However, since 1985, the program has
experienced significant increases in program acquisition unit cost. 
Funding reductions have caused the program to undergo significant
restructuring, resulting in sharp decreases in planned acquisition
quantities and lengthening of development schedules, thereby
increasing Comanche program costs. 

In 1985, the Comanche had estimated total program acquisition costs
of about $61 billion for 5,023 aircraft (or $12.1 million per
aircraft).  In 1992, we reported that (1) as of October 1991, the
program acquisition unit cost had increased to $27.4 million, (2)
acquisition quantities had been reduced to 1,292 aircraft, and (3)
future increases in cost per aircraft were likely.\3 As of February
1995, the Comanche's estimated program acquisition unit cost was
$34.4 million per aircraft, a 185-percent increase from the 1985
estimate.  The estimated total program acquisition cost for the
planned acquisition of 1,292 aircraft is now more than $44 billion. 

Both the Secretary's decision and the Army's restructure would extend
the development program by about 3 years and, under either, increase
the risk of higher total program cost and cost per aircraft. 
However, in reviewing the Army's restructure proposal, DOD noted some
concern over Comanche program costs for fiscal year 2002 and beyond
and the large increase in investment programs projected to occur
about that time.  We are also concerned that the Army's plan to defer
additional developmental efforts to fiscal year 2002 and beyond may
increase the risk that needed funds may not be available to perform
the deferred developmental effort. 


--------------------
\3 Comanche Helicopter:  Program Needs Reassessment Due to Increased
Unit Cost and Other Factors (GAO/NSIAD-92-204, May 27, 1992). 


   TECHNICAL RISKS TO BE RESOLVED
------------------------------------------------------------ Letter :4

The Comanche program's uncertainties in software development and
aircraft maintainability increase the risk that the aircraft will not
perform successfully.  We believe the restructuring provides
additional time to resolve these issues before the decision to enter
production is made. 


      SOFTWARE DEVELOPMENT IS AT
      RISK
---------------------------------------------------------- Letter :4.1

The Comanche will be the most computerized, software-intensive Army
helicopter ever built.  The Army estimates that about 1.4 million
lines of code are required to perform and integrate mission critical
functions.  With additional ground support and training software to
be developed, the total program will have more than 2.7 million lines
of code.  This compares to about 573,000 lines of code for the
upgraded Apache attack helicopter with fire control radar.  The Army
estimates 95 percent of the Comanche's total software will be written
in Ada, a DOD-developed programming language.  The Army plans to
demonstrate initial software performance with the mission equipment
package, which includes the flight control system, during first
flight of the Comanche, scheduled for November 1995. 

The development and integration of on-board, embedded computer
systems is a significant program objective.  The Comanche's
performance and capability depend heavily on these systems and
efforts have been ongoing to solve the problems associated with these
systems.  Nevertheless, (1) software development problems still exist
with the Ada compilation system, (2) delays in software development
and testing are occurring, and (3) improvements are needed in
configuration management.  If these issues are not resolved, the
aircraft's performance and capability will be degraded and first
flight could be delayed. 


         DEVELOPMENT PROBLEMS
         REMAIN UNRESOLVED FOR THE
         ADA COMPILATION SYSTEM
-------------------------------------------------------- Letter :4.1.1

Almost all of the Comanche software will be developed in the Ada
programming language; however, software developers are not using the
same version of the Ada compilation system.  The Ada compilation
system translates Ada code into machine language so that software can
be used by the Comanche's computers.  For example, it is being used
to help develop software for use on the mission equipment package
that is critical for first flight. 

Subcontractors and the contractor team should be using the same,
qualified version of this compilation system to ensure effective
software integration.  However, fixes to individual compiler software
problems are not being shared with all developers; therefore, they
are not using a common compilation system.  These problems have
already delayed qualification testing of the compilation system by 1
year. 

The lack of a uniform, qualified compilation system among software
developers could put first flight at risk, according to the Defense
Plant Representative Office.  Problems with software integration may
show up once integration testing begins in the June to November 1995
time frame.  If that occurs, there may not be time to fix problems
prior to scheduled first flight. 


         DELAYS IN SOFTWARE
         DEVELOPMENT AND TESTING
-------------------------------------------------------- Letter :4.1.2

The program is experiencing high turnover of software engineers at
one of the contractor team's facilities.  In its December 1994,
monthly assessment report, the Defense Plant Representative Office,
which is responsible for contract oversight, observed that high
turnover of software personnel was putting scheduled first flight at
risk.  Loss of key personnel has already contributed to schedule
slippage in several critical software development areas.  Software
development for the following areas has been affected:  the airborne
engine monitoring system, aircraft systems management, control
database, and crewstation interface management. 

The contractor team has formulated a "get well" plan that is
dependent on being able to hire additional personnel in these areas. 
However, hiring additional qualified personnel is difficult,
according to the Defense Plant Representative Office, because
employment would be short term. 

The flight control system software verification testing is also being
delayed.  As of February 8, 1995, Boeing had conducted only 163 of
approximately 500 tests originally planned to be completed by that
date.  The subcontractor responsible for developing this software has
been late delivering software for testing and has provided faulty
software to Boeing, according to the Defense Plant Representative
Office.  Boeing established a recovery plan for this area that would
have resulted in a completion date in March 1995--about a 1-month
delay from the original plan.  However, in February 1995, the
contractor revised the recovery plan to reflect a completion date of
July 1995--a 5-month delay. 

The flight control system is critical to first flight, according to
the Defense Plant Representative Office.  However, because of delays
with verification testing, the Defense Plant Representative Office is
concerned that the remaining verification testing, as well as, the
validation and formal qualification testing will not be completed in
a timely manner.  As a result, first flight may be delayed.  Boeing
is scheduled to complete these tests prior to first flight. 
According to the program office, Boeing's plan to complete the
testing calls for it to be conducted concurrently.  If major problems
occur in any one of the testing phases, there may not be enough time
to fix the problem and complete all testing before first flight. 


         IMPROVEMENTS IN
         CONFIGURATION MANAGEMENT
         ARE NEEDED
-------------------------------------------------------- Letter :4.1.3

Configuration management is the discipline of applying technical and
administrative direction and surveillance to (a) control the flow of
information between organizations and activities within a project;
(b) manage the ownership of, and changes to, controlled information;
(c) ensure information consistency; and (d) enable product release,
acceptance, and maintenance.  The part of configuration management
used to report software problems and changes among the contractor
team and subcontractors has shortcomings that put software
development at risk. 

In its November 1994 monthly assessment report, the Defense Plant
Representative Office observed that the lack of a common problem
reporting system made proper handling of software related changes
difficult.  Furthermore, the report noted that this situation could
adversely impact scheduled first flight of the Comanche.  As of
February 1995, the contractor team still did not have a common,
automated database available to track problem change reports.  Thus,
the contractor team, as well as subcontractors, did not have
visibility over changes made to software. 


      KEY MAINTAINABILITY
      REQUIREMENTS AT RISK
---------------------------------------------------------- Letter :4.2

Maintainability requirements are important to achieving lower
operating and support costs and wartime availability goals.  However,
these goals are at risk because key maintainability requirements such
as direct maintenance man-hours per flight hour (MMH/FH), the mean
time to repair (MTTR), and fault isolation may not be achievable. 
Individually, failure to meet these parameters may not be a
significant problem; however, collectively they affect the ability of
the Comanche to achieve lower operating and support cost and wartime
availability objectives. 


         KEY MAINTENANCE GOAL MAY
         NOT BE REALISTIC
-------------------------------------------------------- Letter :4.2.1

In March 1987, the Army established a 2.6 direct MMH/FH requirement
for the Comanche.  It represents the corrective and preventive
maintenance per flight hour expected to be performed at the unit
level.  The Army formulated its planned wartime operating tempo for a
Comanche battalion based on 6 hours a day per aircraft, or 2,200
flying hours per year.  It then determined the maintenance factor
needed to support this operating tempo--2.6 MMH/FH.  As the MMH/FH
level increases, the number of maintainers needed to sustain the
2,200 wartime flying hour goal increases, as do operating and support
costs.  Conversely, if the Army could not increase the number of
maintainers, the planned operating tempo would have to be reduced. 

The reasonableness of the Comanche's 2.6 direct MMH/FH requirement
has been debated for several years within the Army and DOD. 
Representatives from the program office; the Army Materiel Systems
Analysis Activity, which independently evaluates program testing
results; the Office of the Assistant Secretary of the Army for
Research, Development, and Acquisition; and the Army Cost and
Economic Analysis Center met on October 28, 1994, to discuss the
direct MMH/FH goal for the Comanche program.  They agreed that the
2.6-MMH/FH requirement was not a realistic, achievable goal. 
Consequently, Army officials reached consensus and agreed on 3.2
direct MMH/FH as the Army-wide position for this parameter.  However,
during these discussions, Army Materiel Systems Analysis Activity
personnel noted that attaining a 3.2-MMH/FH goal represented a medium
to high risk, while a 4.3-MMH/FH goal had a low to medium risk. 
Increasing the maintenance factor increased the number of maintainers
needed and will increase estimated operating and support costs by
about $800 million over a 20-year period. 

The direct MMH/FH requirement does not represent the total
maintenance burden for the Comanche because it does not include
indirect maintenance time.  The Army does not normally collect data
on indirect maintenance time.  According to the program office, its
best estimate of indirect maintenance time, following Army guidance,
is 2.5 MMH/FH, and this figure has been used for calculating manpower
needs for crew chief personnel on the Comanche.  Thus, the total
maintenance burden assumed for the Comanche is currently 5.7 MMH/FH
(3.2 direct MMH/FH plus
2.5 indirect MMH/FH). 


         REPAIR REQUIREMENT MAY
         NOT BE MET
-------------------------------------------------------- Letter :4.2.2

To minimize turnaround time for repairs at the unit and depot, the
Army established MTTR requirements of 52 minutes for repairs at the
unit level and up to 12 hours at the depot level for the Comanche. 
These requirements represents the average time expected to diagnose a
fault, remove and repair an item, and perform an operational check
and/or test flight.  We determined that any increase in MTTR above 1
hour will begin to impact the Army's wartime availability goal of
2,200 hours per year, unless additional maintenance personnel are
available. 

As of January 1995, the contractor team was estimating that the Army
would achieve 59 minutes for unit level repairs.  According to
contractor team officials, the requirement was not being met because
the cure time required for composite material used on the aircraft
was greater than expected.  The contractor team discussed changing
the MTTR requirement to 1 hour; however, the program office believes
the problem could be resolved and did not believe the specification
should be changed.  The contractor team has not yet developed MTTR
estimates for depot-level repair. 


         KEY DIAGNOSTIC SYSTEM
         REQUIREMENT MAY NOT BE
         ACHIEVED
-------------------------------------------------------- Letter :4.2.3

The Comanche's diagnostic system is required to correctly isolate
failed mechanical and electrical components at least 80 percent of
the time with a high degree of accuracy.  A high level of accuracy is
essential as it allows maintainers to isolate and fix problems at the
unit level.  If the fault isolation requirement is not met, the
Comanche is unlikely to achieve its MTTR requirement, thereby
adversely affecting the Army's ability to execute its maintenance
concept and its wartime availability goals. 

Contractor team officials stated the fault isolation requirement was
very optimistic, and although they are striving to meet this
requirement, it may eventually have to be changed.  As of January
1995, the contractor team predicted the system could achieve an
overall 69-percent fault isolation rate; however, this rate would not
meet the specification for mechanical and electrical component fault
isolation.  There are design limitations on two components, according
to the program office, and changes to bring these components into
conformance with specifications would be costly and increase the
weight of the aircraft.  Therefore, as of January 1995, the
contractor team and the program office have agreed not to take action
on these components. 


         FALSE REMOVAL RATE IS
         HIGH RISK
-------------------------------------------------------- Letter :4.2.4

The Army established a requirement of a 1-percent false removal rate
for the Comanche.  A false removal occurs when a part removed from
the aircraft shows no evidence of failure when tested.  This
requirement is dependent, to a large extent, on the success of the
fault detection/isolation system in detecting and isolating failed
components.  Program personnel characterize the 1-percent requirement
as stringent and one that will be challenging to achieve.  An Army
Materiel Systems Analysis Activity official believes some design
improvements have occurred in this area, but the risk associated with
achieving this requirement still remains high.  If the Comanche does
not meet this requirement, estimated operating and support costs for
the Comanche will be higher than previously predicted. 

The Army has not had good experience in developing fault
detection/isolation and false removal systems for other aircraft.  In
September 1990, we reported that the fault detection and isolation
system on the Apache aircraft did not always accurately detect the
component that caused a particular fault, and the system detected
faults that did not actually exist about 40 percent of the time.\4 As
a result, Apache maintainers had to perform additional work to locate
failed components.  Recently, through a reliability program, the
false removal rate for the targeting and night vision systems on the
Apache improved to about 10 to 15 percent, according to Army
officials.  This is still significantly higher than the 1-percent
requirement established for the Comanche program. 


--------------------
\4 Apache Helicopter:  Serious Logistical Support Problems Must Be
Solved to Realize Combat Potential (GAO/NSIAD-90-294, Sept.  28,
1990). 


   SOME PROGRAM GOALS ARE
   CURRENTLY BEING MET
------------------------------------------------------------ Letter :5

Although the program is experiencing technical problems, it is
currently meeting its goals of reducing maintenance levels and
keeping overall weight growth within acceptable limits for the
Comanche. 

The Army's maintenance concept for the Comanche program is predicated
on two levels of maintenance--unit- and depot-level maintenance. 
This concept is important to achieving operating and support savings
predicted for the program because it eliminates the intermediate
level of maintenance.  Unit-level maintenance entails removing and
replacing components required to return the aircraft to a serviceable
condition.  Depot-level maintenance requires higher level maintenance
skills and sophisticated capital equipment and facilities not found
at the unit level.  The Army traditionally uses a three-level
maintenance concept that includes intermediate-level maintenance to
handle component repairs.  Intermediate-level maintenance is usually
located close to the battalion.  It is performed on components that
cannot be easily repaired at the unit level and do not require the
more sophisticated repairs done at the depot level. 

As of January 1995, no Comanche component had been designated for
repair at the intermediate level, according to the program office. 
Contractor team personnel are conducting repair level analysis on
Comanche components to determine whether components should be
repaired at unit, intermediate, or depot facilities, according to
program and contractor team officials.  Any candidates identified for
intermediate-level repair are reviewed for possible design changes
that could allow maintenance at the unit or depot level.  If
economically feasible, the contractor team will make design changes
to the component to preclude the need for intermediate-level repair. 

As of February 7, 1995, the Comanche's empty weight increased from
its original specification of 7,500 pounds to 7,883 pounds.  Although
the Comanche's weight continues to increase, it remains within the
allowable design limit of 7,997 pounds.  Weight increases affect
vertical rate of climb performance on the Comanche.  The Army
established a limit of
500 feet-per-minute as the minimum acceptable vertical rate of climb
performance.\5 If the Comanche's weight exceeds 8,231 pounds, the
engine will have to be redesigned to produce enough power at 95
percent maximum rated engine power to sustain the minimum 500
feet-per-minute vertical rate of climb requirement. 


--------------------
\5 At 95-percent maximum rated engine power. 


   RECOMMENDATION
------------------------------------------------------------ Letter :6

We recommend that the Secretary of Defense require the Army to
complete operational testing to validate the Comanche's operational
effectiveness and suitability before committing any funds to acquire
long-lead production items or enter low-rate initial production. 


   AGENCY COMMENTS AND OUR
   EVALUATION
------------------------------------------------------------ Letter :7

DOD generally concurred with the findings and original
recommendations in our draft report.  In commenting on the draft
report, DOD offered explanations about why the problems that we
identified were occurring and what they were doing to fix those
problems.  DOD disagreed with the report's conclusion about false
removals and stated that we had not presented any evidence that the
Comanche's 1-percent false removal rate may not be achievable.  We
still believe that the false removal goal is high risk and adjusted
the report to more clearly reflect our concern. 

Regarding our draft report recommendation that DOD develop program
fixes that achieve program goals and reduce the risks we identified,
DOD concurred and noted that the approved restructuring will
significantly reduce risk.  DOD concurred with our other draft
recommendation not to commit production funds to the program until
performance and mission requirements are met and noted that the
program would be reviewed by DOD before approving the Army's request
to proceed to the engineering and manufacturing development
phase--the Milestone II decision scheduled for October 2001. 

Because DOD concurred in our draft report recommendations and is
taking action on them, we are no longer including them in this
report.  However, our analysis of information on the restructuring
obtained after we had submitted our draft report to DOD has further
heightened our concerns about the risk of concurrency; therefore, we
have revised the report and added a new recommendation.  Under the
stretched out, restructured Comanche program, operational testing is
not even scheduled to begin until after the low-rate initial
production decision is made.  This approach continues the risks
associated with making production decisions before knowing whether
the aircraft will be able to perform as required. 

Prior to the restructure, the Army planned to start operational
testing with eight aircraft in May 2003.  Under the restructured
program, the Army plans to start operational testing with six
helicopters by about August 2005.  We believe that the stretched out
time frame and the six aircraft acquired under the restructure
provide sufficient time and aircraft to operationally test the
Comanche prior to making any production decisions. 

Additionally, because operational testing is not scheduled until
about August 2005, DOD will not be in a position at Milestone II in
October 2001 to adequately address whether the Comanche program is
meeting its performance requirements.  DOD's comments are presented
in their entirety in appendix I, along with our evaluation. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :8

To assess cost changes, software development, maintainability, and
weight growth issues, we reviewed program documents and interviewed
officials from the Department of the Army headquarters, Washington,
D.C.; the Comanche Program Manager's Office, St.  Louis, Missouri;
the U.S.  Army Materiel Systems Analysis Activity, Aberdeen Proving
Ground, Maryland; the Ada Validation Facility, Wright-Patterson Air
Force Base, Ohio; and the Office of the Assistant Secretary of
Defense for Program Analysis and Evaluation, Washington, D.C.  We
also reviewed program documents and interviewed contractor and
Defense Plant Representative Office officials at the Boeing
Helicopter Company, Philadelphia, Pennsylvania; the Sikorsky Aircraft
Corporation, Stratford, Connecticut; and the Comanche Joint Program
Office, Trumbull, Connecticut. 

We conducted our review between August 1994 and February 1995 in
accordance with generally accepted government auditing standards. 


---------------------------------------------------------- Letter :8.1

We are also sending copies of this report to the Chairmen and Ranking
Minority Members of the Senate and House Committees on
Appropriations, the Senate Committee on Governmental Affairs, and the
House Committee on Government Reform and Oversight; the Director,
Office of Management and Budget; and the Secretaries of Defense and
the Army.  We will also provide copies to others upon request. 

This report was prepared under the direction of Thomas J.  Schulz,
Associate Director, Systems Development and Production Issues. 
Please contact Mr.  Schulz at (202) 512-4841 if you or your staff
have any questions concerning this report.  Other major contributors
to this report are listed in appendix II. 

Louis J.  Rodrigues
Director, Systems Development and
 Production Issues




(See figure in printed edition.)Appendix I
COMMENTS FROM THE DEPARTMENT OF
DEFENSE
============================================================== Letter 

See comment 1. 

See comment 2. 



(See figure in printed edition.)



(See figure in printed edition.)

Now on pp.  3-5. 

See comment 1. 



(See figure in printed edition.)

See pp.  11-12. 

Now on p.  5. 



(See figure in printed edition.)

Now on pp.  5-6. 



(See figure in printed edition.)

Now on pp.  6-7. 

Now on p.  7. 



(See figure in printed edition.)

Now on pp.  7-8. 

See comment 3. 



(See figure in printed edition.)

Now on pp.  8-9. 

See comment 3. 

Now on p.  9. 



(See figure in printed edition.)

See comment 3. 

Now on pp.  9-10. 

See comment 4. 



(See figure in printed edition.)

Now on pp.  10-11. 



(See figure in printed edition.)



(See figure in printed edition.)

Now on p.  12. 

See pp.  11-12. 

Now on p.  12. 

See pp.  11-12. 


The following are GAO's comments on the Department of Defense's (DOD)
letter dated April 20, 1995. 

1. As DOD's comments note, there are many measures of unit cost, such
as average unit flyaway cost, program acquisition unit cost, and unit
procurement cost.  We believe that the program unit cost that we used
in the report--which the footnote in the report defines as total
research, development, and acquisition costs in current dollars--is
as valid as flyaway cost to portray program cost growth over time. 
We have adjusted the report to more clearly define the basis of the
unit cost we use. 

2. These comments are dealt with on pages 11 and 12 of the report and
in our responses to the specific DOD comments that follow.  Report
material on costs and concurrency has been revised to reflect
information obtained after our fieldwork had been concluded. 

3. The report does not say that maintainability goals will never be
met.  We pointed out that some key maintainability requirements are
not being met and, therefore, there is a risk that the Army may not
achieve the lower operating and support costs and wartime
availability goals that it has established for this program.  We also
said that individually, failure to meet these parameters may not be a
significant problem; however, collectively they affect the ability of
the Comanche to achieve the cost and availability goals.  This point
is clearly illustrated in DOD's comments on the failure of the fault
isolation system.  According to DOD, "Fault isolation is one of the
key diagnostic system requirements.  The DOD agrees that if the fault
isolation requirement is not met, the Comanche is unlikely to achieve
its mean-time-to-repair requirement, .  .  .". 

4. We still believe that this goal is very aggressive.  DOD
acknowledges that this goal is stringent and the Army has not had
good experience in the past with false removals on other aircraft. 
Additionally, as noted in the report, Army Materiel Systems Analysis
Activity said the risk associated with achieving this requirement
remains high.  We changed the section heading to emphasize the high
risk. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix II

NATIONAL SECURITY AND
INTERNATIONAL AFFAIRS DIVISION,
WASHINGTON, D.C. 

Robert J.  Stolba

ACCOUNTING AND INFORMATION
MANAGEMENT DIVISION

David Chao

KANSAS CITY REGIONAL OFFICE

Gary L.  Billen
Robert D.  Spence
Lauri A.  Bischof
Michael W.  Buell
Karen A.  Rieger
