I Jrritwl St,;O.cs (~errc?

ral Accounting

GA6

Offic~ : --

- Report, to t/he Acting Secretary of the

Navy

ADP PROCUREMENT
Prompt Navy Action
Can Reduce Risks to
SNAP III
Implementation

GAO

United States
General Accounting OfRce
Washington, D.C. 20548
Iniormation Management and
Technology Division

B-260036
September 29, 1992
The Honorable Sean C. OKeefe
The Acting Secretary of the Navy
Dear Mr. Secretary:
This report presents the results of our risk assessment of the Navys
efforts to upgrade its Shipboard Non-Tactical ADP Program (SNAP). We
identified potential acquisition risks and tested our recently developed risk
assessment methodology, which incorporates critical factors identified in
our model of the information technology acquisition process. The
methodology is designed to provide an early warning to agency
management so they can take timely action to address potential risks.
We selected this acquisition, known as SNAP III, because of its large
life-cycle cost, estimated at about $600 million,2 and because it is in an
early stage in the procurement cycle, where corrections are easier and
cheaper to make. Our review focused on the role of senior managers in
system development and the Navys plans for implementing SNAP III.
Details of our objectives, scope, and methodology are discussed in
appendix I.

Results in Brief

The Navys current management structure does not adequately support the
SNAP III project. Although SNAP III has a program manager and a program
sponsor, neither has authority or control over the development and
funding for the SNAP III system. Instead, 12 different commands have
authority over the development, funding, and procurement of the major
systems that comprise SNAP III. Without a central authority, the Navy
cannot prevent multiple, incompatible hardware and software systems
from being developed and installed. These redundant development efforts
require excessive financial resources and increase the amount of training
and maintenance needed to support the different systems.
The Navy also lacks a strategic information resources management (IRM)
plan for SNAP III. Such a plan is critical to a projects success because it
identifies and defines key characteristics, such as the systems hardware,
Information Technology:A Modelto Help ManagersDecreaseAcquisition Risks(GAO/IMTEC8.1.6,
Aug. 1990).
2SNAPIII life-cyclecosts are subjectto changependingthe outcomeof the Navysdownsizingefforts,
which include the decommissioningof an unknown numberof ships.

GAO/IMTEC-9249 ADP Procurement

Page 1

,,

software, and interface standards, and presents a structured approach for

achieving them. Instead, the Navy is addressing these key characteristics
on an ad hoc basis. For instance, the Navy was planning to acquire SNAPIII
nontactical hardware from a future tactical hardware contract without
first defining the required hardware performance and capacity or
preparing a contingency plan in case this procurement cannot be used.
The Navy is also developing upgraded software that does not adhere to
federal standards and is continuing to install old and incompatible systems
aboard the Navys newest ships.
On the basis of our findings, the Vice-Chief of Naval Operations and the
Chief of Naval Operations (CNO) Chief Information Officer agreed to
(1) consolidate all shipboard nontactical ADP programs under a single
command with authority and control over the funding and development,
and (2) develop a strategic plan for all shipboard nontactical ADP
programs.
The SNAP mission is to reduce the administrative burden on sailors, and
includes all afloat and shore-based automated initiatives. SNAP consists of
six major automated systems that support the maintenance,
supply/financial, aviation, pay/personnel, administration, and
medical/dental functions,

Background

The SNAPprogram was originally implemented as two distinct systems,

both installed in the early 1980s.SNAP I, using Honeywell DPS-6 computers,
was installed aboard the Navys largest vessels, such as aircraft carriers
and tenders. SNAP II, which uses Harris H-300 computers, was installed on
smaller surface vessels and submarines. Although both systems perform
the same functions, they were developed independently, have different
operating systems, and are not compatible with one another.
SNAP I and SNAP II are characterized by

high training and maintenance costs,

. saturated or limited capacity,
proprietary operating systems, and
obsolete 1970s hardware.
l

Recognizing the limited capacity of both systems, as well as the

unnecessary training and maintenance needs they create, the Navy
initiated a replacement effort--SNAP III. SNAP III is to be a unified system

Page 2

GAO/IMTEC-9249 ADP Procurement

1,

B-269035

that will incorporate all nontactical shipboard systems, allow users to

access the functional applications from any terminal, use a standard
operating system, and interconnect via the ships computer network. SNAP
III goals include
reduced training and maintenance costs,
increased system access,
elimination of redundant applications software and multiple operating
systems,
the use of open systems, and
the use of fiber optics.
The concept development phase of SNAP III is scheduled for completion in
September 1992, with deployment scheduled to start in 1994.

Autonomous
Commands Prevent
Effective
Development Of
SNAP III

Our model and audit methodology stress that to be effective, agencies

should be organized in such a way that a single manager (1) has complete
authority over all phases of the acquisition process and (2) is responsible
and accountable for ensuring that the acquisition is effectively and
efficiently conducted. The managers responsibility should include
authority over program funds and implementation of the system.
While several top-level Navy officials are involved with SNAP III, no single
manager or organization is directing the development or controlling SNAP'S
deployment, As a result, the Navy has redundant development efforts in
process. These multiple efforts result in incompatible systems that require
excessive funding and an increased amount of training and maintenance.
Numerous commands are involved with the SNAP program. The Space and
Electronic Warfare Directorate is SNAP'Sprogram sponsor, while the Space
and Naval Warfare Systems Command (SPAWAR) is the SNAP program
manager. Further, 12 separate commands serve as either functional
sponsors or managers for SNAP'Ssix major systems.
The program sponsor serves as an advocate for system development. A
significant responsibility of the program sponsor is to ensure that funds
are available to support SNAP. In addition, the program sponsor is
responsible for coordinating with the functional sponsors, functional
managers, and fleet users to ensure that SNAP systems adhere to the Navys
objectives, priorities, functional requirements, and the anticipated
operating environment.

Pq3e 3

GAOAMTEC-92-69ADP Proeurement

B-269096

SPAWAR,as program manager, is responsible for planning and managing the

overall SNAPprogram. Its overall responsibility is ensuring that SNAP

complies with Navy standards and meets usersneeds. The program

manager is also responsible for establishing and maintaining configuration
control procedures for SNAP hardware and software, and coordinating the
installation of SNAP hardware on Navy ships. SPAWARalso has a software
development group that is responsible for maintaining and enhancing
existing sNApsoftware.
While both the SNAPprogram sponsor and program manager have critical
responsibilities, neither has the authority to ensure that these
responsibilities are met. Instead, the 12 commands in which the six
functional sponsors and six functional managers reside have authority
over systems supporting their area of responsibility and control both
system development and funding. Consequently, 12 separate commands
have autonomy over the development and funding of SNAP'Smajor systems.
The following examples show how a lack of a central authority over
programs and their funding can result in multiple, inefficient systems.

Development of Multiple
MaintenanceSystems

While SPAWARSresponsibilities include maintaining and enhancing the SNAP

maintenance system, SPAWARcannot (1) prevent other maintenance
systems from being developed and implemented outside of the SNAP
program, or (2) control the types of computers that are installed to
support these systems.
Because the maintenance functional manager, the Naval Sea Systems
Command (NAVY@, was not satisfied with the existing SNAP maintenance
system, it used its autonomy to spend about $70 million (through the end
of calendar year 1991) developing its own maintenance system. NAVSEA
procured new computers, developed new maintenance software, and
lntended to connect its system to SNAPSelectronically. NAVSEAdeveloped
its system during the same period that SPAWARupgraded the existing SNAP
maintenance software. As a result, the Navy ended up with two
independently developed systems that replicate functions and are
incompatible with one another.
On March 20,1992, SPAWARsent a memorandum to NAVSEAstating that
because the NAVSEAsystem interface was not in compliance with SPAWARS
military standard (MILSTD-461) required for connection to the SNAP

Page 4

GAGAMTEC-9249 ADP Procurement

B-250986

system, it should not be installed until this standard was met.3 However, on
March 23,1992, NAVSEZA
had SPAWAR'supgraded maintenance software
removed from the SNAP system on an Atlantic Fleet tender and installed its
own computers to replace the sNApmaintenance function.
Because the NAVSEAsystem did not meet the SNAP interface standard, the
two systems (which resided in the same room) could not be connected
electronically. Instead, data had to be batched on magnetic tape and hand
carried between the NAVSEAand SNAP computers. This manual processing
increased the time required to process material requisitions from 10
seconds using SPAWAR'Supgraded SNAP maintenance system to 48 hours
using the NAVSEAsystem. As of the end of August 1992, the interface
problem had not been solved and the Navy was continuing to use two
inefficient systems to support its SNAPfunctions.
Moreover, NAVSEAinstalled three different hardware and software versions
of its maintenance system in the Pacific and Atlantic Fleets. This occurred
because various commands insisted that the system be installed on
computers that met their individual needs.
This redundant approach to maintenance automation is contrary to SNAP
IIIs goal of developing a standard system on all Navy ships. However, the
program coordinator informed us that the NAVSEAsystem will be installed
on several additional ships because funds are already obligated. He added
that the strategic IRM plan for SNAP III will address how the NAVSEA
maintenance system will transition to the new environment.

Development of an
Incompatible System

Another project that demonstrates the risks inherent in the Navys

decentralized management structure is Micro-SNAPII. Initiated in 1986,
Micro-SNAPII is an effort to automate small ships and shore sites using
microcomputers and SNAP II functions. Micro-SNAPII is not compatible with
SNAP III'S go&S.

Micro-SNAPII is being designed using Microsofts disk operating system,

which is not compatible with the UNIX open system standard the Navy
selected for SNAP III. Also, Micro-SNAPII is not being installed with fiber
optics and does not adhere to approved Navy security standards. The SNAP
program sponsor told us the software used for Micro-SNAPII would be
scrapped when SNAP III is implemented, but that the hardware would be
!'MILSTD461defineselectromagneticemissionand susceptibilityrequirementsfor the control of
electromagneticinterference.
Pnge5

GAO/IMTEC-9249ADPProcurement

&

B.2110026

retained. Other Navy officials characterized Micro-sNAp II software as a

throw-away system. Thus, in a period of declining resources, the Navy is
simultaneously developing Micro-sNApII and the system that will
replace it.
At the end of August 1992, the SNAP program manager said the Navy
planned to continue with the development and eventual installation of
Micro-sNApII systems. For example, he said his office is in the process of
obtaining security accreditation of the Micro-SNAPII system. However, he
did not anticipate receiving this accreditation for at least 6 months. He
added that fiber optics would be used on future installations of the
Micro-SNAPII system. He also said the systems are needed aboard ships not
currently automated and that ongoing planning efforts will describe how
the system will transition to the SNAP III environment.

Absence of Strategic
Plan Prevents an
Efficient Approach to
SNAP Development
and Implementation

Our model reflects the importance of a strategic IRM plan to help agencies
define and implement systems that meet their needs. The plan should
describe the hardware, software, and communications that will be needed
to meet the mission most cost-effectively. In addition, a plan should
identify how management will transition from old to new systems.
Implementation of the Navys SNAP III project reflects the lack of a
strategic IRM plan. For example, the SNAP III project manager planned to
use an existing tactical procurement contract to obtain nontactical SNAP
hardware. However, SPAWARdid not develop a contingency plan or identity
hardware capacity needs--factors that would be addressed in a strategic
IRM plan. SPAWARjustified this approach by stating that (1) the existing
hardware was extremely antiquated; any new hardware would be better
than the existing hardware, and (2) senior management officials had
planned to obtain a waiver from federal regulations in order to obtain
nontactical hardware from a tactical Navy contract.
SPAWARhad not considered the possibility that senior Navy officials would

be unable to obtain a waiver from federal regulations and thus did not
develop any contingency plans for SNAP IIIs hardware acquisition. By not
establishing alternative acquisition approaches, the Navy is at risk of
significantly delaying the SNAP III project.
In addition, no plan has been developed to ensure that compatible
software is provided for all nontactical systems. Without a plan, the Navys
Central Design Activity for nontactical systems is continuing to develop

Page 6

GAOAMTEC-92-69ADP Procurement

B.260086

and maintain incompatible SNAPsoftware for more than four different SNAP
operating systems.
Further, because the Navy has not established communications standards,
multiple, incompatible networks are being installed. Some ships have two
or more networks that are incompatible. Also, while the Navy has
established a SNAP III communications goal of equipping all ships with fiber
optics, it has no plan showing how this will be achieved.
Without a strategic IRM plan, the Navy continues to install antiquated SNAP
II systems on its newest vessels. Under this approach, the Navy does not
know whether the additional costs that will be incurred to remove this old
equipment and replace it with the new SNAP III system are justified.
At our suggestion, the Navy has begun to develop a strategic IRM plan for
meeting its SNAP III goals. According to the program sponsor, the strategic
IRM plan will address the issues we raised regarding software,
communications, and the transition from old to new systems. He also told
us that the Navy would request a delegation of procurement authority
(DPA) from the General Services Administration to obtain SNAP III
hardware. However, as of the end of August 1992, a DPA had not been
requested. In addition, the program manager told us in August 1992 that
the Navy had begun to identify the SNAP III hardware capacity needs and
reach agreement on the communications standard to be used for all
shipboard networks.

Conclusions

The success of information system acquisitions can be measured by their

cost effectiveness in assisting agencies in carrying out their missions and
by the timeliness of delivery. Success with the SNAP III acquisition requires
that the Navy act responsibly to manage and control acquisition risks.
The Navy is not currently providing sufficient managerial oversight to
minimize SNAP III risks. This has resulted in incompatible systems, which
require different hardware and software to perform similar functions,
being installed aboard ships. Continued implementation of incompatible
systems is contrary to the Navys goal of creating an integrated
environment for all shipboard nontactical systems. A single authority over
the development and funding of all shipboard nontactical programs is
needed to ensure that all ongoing and planned efforts support this goal.

GAOIIMTEC-92-69ADP Procurement

Page 7

Furthermore, the Navy lacks a strategic IRM plan for SNAP III
implementation. Without a plan specifying standards for hardware,
software, and system interfaces, all efforts to integrate the shipboard
envlronment will be at high risk for schedule delays, cost overruns, and
poor performance. Lacking a strategic IRM plan with a detailed transition
approach, the Navy cannot determine whether continued installation of
old systems on new ships is justified.

Recommendations

In order to minimize SNAP III risks, we recommend that the Acting

Secretary of the Navy suspend all nontactical ADP development until the
Navy completes
restructuring of the management over shipboard nontactical programs to
provide a single manager with authority over funding and development for
all shipboard nontactical ADP programs, and
development of a strategic IRM plan for all shipboard nontactical ADP
programs which describes the transition from the current environment to
SNAP III. This plan should include sections which identify hardware
capacity needs; contain alternative technical and procurement
approaches; and include standards for hardware, software, and shipboard
networks.

Agency Comments
and GAOsEvaluation

Navy officials agreed in general with our recommendations and stated they
are taking action to address them. For example, the Vice-Chief of Naval
Operations has initiated action to provide a single authority over all
shipboard nontactical ADP initiatives. The Navys action designating the
Director, Space and Command, Control, Communications and Computer
Systems Requirements as the single resource sponsor should significantly
decrease the risks associated with attempts to provide an integrated
system on Navy ships. Finalizing and implementing a strategic IRM plan will
also help the Navy achieve its SNAP III goals.
These actions by themselves, however, do not guarantee an efficient and
effective system. The Navy must carefully manage SNAP III throughout its
life cycle to ensure it is implemented successfully.
As you know, 31 U.S.C. 720 requires the head of a federal agency to submit
a written statement on actions taken on our recommendations to the
Senate Committee on Governmental Affairs and the House Committee on

Page 9

GAO/IMTEC-92-69

ADP Procurement

B-260026

Government Operations not later than 60 days after the date of this letter.
A written statement must also be submitted to the House and Senate
Committees on Appropriations with the agencys first request for
appropriations made more than 66 days after the date of this letter.
We are sendlng copies of this report to the Chairmen of the Senate and
House Committees on Appropriations; the Chairmen of the Senate and
House Committees on Armed Services; the Director, Office of
Management and Budget; and the Secretary of Defense. Copies also will be
made available to other interested parties upon request.
We performed our work in accordance with generally accepted
government auditing standards from October 1991 through August 1992, in
Washington, D.C., and Norfolk, Virginia This work was performed under
the direction of Jack L. Brock, Jr., Director, Governmentwide Information
and Financial Management, who can be reached at (202) 5124406. Other
major contributors to this report are listed in appendix II.

Ralph V.-Carlone
Assistant Comptroller General

GAO/IMTJX-92-69 ADP Procurement

Page 9

Anoendix I

Objectives, Scope, and Methodology

We reviewed the Navys Shipboard Non-Tactical ADP Program (SNAP)

replacement effort as part of our responsibility to ensure that federal
agencies acquire and use information technologies in a cost effective
manner. Our objectives were to (1) identify potential risks related to the
SNAP III systems development, and (2) test a risk assessment methodology
we are developing. We chose to review the SNAP project because it had a
projected cost of more than $600 million over the systems life cycle, and
the project is currently in the concept development phase, where
suggested improvements can be made more easily and cheaply.
Our review focused on issues relating to the upgrade of existing SNAP I and
II systems and the role of Navy managers in planning and
implementing a new shipboard computer environment. We used our
acquisition audit methodology and GAO'S acquisition model to assess
whether the project had any maor risks that should be addressed by Navy
management.
SNAP

We interviewed functional managers and sponsors in Washington, D.C., for

each of SNAPS major systems including the maintenance, supply/financial,
aviation, pay/personnel, administration, and medical/dental functions. We
reviewed key SNAP program briefing documents and transcripts of SNAP III
meetings. We also met with staff from the Navy Audit Service and the
General Services Administrations Office of Information Resources
Management Policy regarding their efforts associated with the SNAP
program. In Norfolk, Virginia, we met with officials at the
Commander-in-Chief, Atlantic Fleet headquarters where we observed
operational SNAP systems on four types of ships-the USS John F. Kennedy
(aircraft carrier CV 67), the USN Baltimore (submarine SSN 704), the USS
Shenandoah (tender AD 44), and the USS Yorktown (cruiser CG 48).

(GAO/IMTEC 8.1.6, Aug. 1990)

Pnge 10

GAo/rMTEC-82-69

ADP Procurement

Appendix II

Major Contributors to This Report

Information
Management and
Technology Division,
Washington, DC.

(610774)

Mark E. Heatwole, Assistant Director

Bernard R. Anderson, Senior Evaluator
Peter C. Wade, Senior Evaluator

Page 11

GMMMTEC-92-69

ADP Procurement

Ordt~ring Informat

ion

ItrtA first copy of each G A O report and testimony is fret*. Additional

copies art $2 each. Orders should be sent. to the following address,
;~t~t~otn~)anit~tiby a check or money order made out t.n)the Superintt~ndtwt of I)ocuments, when necessary. Orders for 100 or more
copiths to be mailtv~ t,o a s ingle address are discounttvi 25 ptArct~nt.
1i.S. (;t~nt~ral Accounting O ffice
I.(). 130x 6015
(Gail trrrsbnrg, MD 20877
Ordtbrs may also be placed by calling

(202) 2754241