Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 15
2
MANAGEMENT CHALLENGES IN MODERNIZING LOGISTICS SYSTEMS
In our midterm report we said that the Logistics Systems
Modernization Program (LSMP) could be the Defense Logistics Agency's
(DLA's) blueprint for excellence by providing the driving force for
deploying and executing its modernization plans. We continue to
believe in that possibility. However, we note that the program has
achieved less than expected during the course of our study.
The LSMP was initiated in 1986 to modernize the agency's logistics
systems by providing for data-sharing and improvements in the utility
of the information provided. It was presented to us as a
technology-based program that would be implemented over a number of
years in a carefully reasoned and executed series of steps. We agreed
with the need for the program and its underlying logic. We remain
supportive of the need for such a program as part of the process by
which the agency achieves its vision of how it will operate in the
future.
We believe that an agency's modernization should be guided by a
unified concept rather than a collection of individual fix-it
programs. In the past, the DLA upgraded its systems on an "as needed"
basis and in a piecemeal fashion. In other cases, beneficial upgrades
often were not made because of budgetary constraints or the absence of
urgency. As a result, over time the agency's systems have become
fragmented and outdated. Hence there are significant opportunities to
improve the efficiency and effectiveness of the agency, and unless its
present mode of operation is corrected, these opportunities will remain
unrealized.
Recently, grand designs that attempt to combine all related-
requirements in a single package have become a subject of debate.
Their very largeness and complexity are said, by some, to be obstacles
to success (U. S. General Services Administration, 1988~. We;are not
opposed to grand designs per se. In some cases they may be the only
way to achieve a grand result. However, we do recognize the risks
associated with grand implementations and have suggested approaches
that will help the agency avoid across-the-board modernization without
abandoning its view of the future. In so doing, we hoped to guide the
development of the agency's modernization plans as the agency itself
evolves toward moving more information and less materiel.
15
OCR for page 16
16
RESPONSIBILITY, AUTHORITY, AND ACCOUNTABILITY
A principal concern of ours has been the perceived lack of clarity
and focus for such management basics as the identification of
responsibility, authority, and accountability for the various actions,
phases and steps associated with systems modernization.
The scale and complexity of the LSMP requires strong management and
organizational cooperation. We have consistently believed that the
effort dictated by the LSMP requires the full attention of a management
group whose sole responsibility is the successful achievement of
modern, interoperable information systems. Furthermore, in order to
accomplish such internal integration, this management group must have
the responsibility, authority, and accountability to enforce program
objectives across the diverse and autonomous organizations and
activities of the DLA. In addition, the interest and attention of top
management to the key initiatives of the program must permeate the
entire organization. The highest levels of the DLA should be active in
a planning process that scopes the LSMP for implementation by the
management group discussed above. An oversight and tracking apparatus
for the LSMP should be institutionalized in such a manner that the key
corporate staff elements are continuously aware of staff disagreements,
program delays, design failures and successes.
Our Midterm View
In our midterm report we supported the-DLA's creation of a program
office and recommended that it be strengthened by adding staff with
managerial and technical skills. We urged the DLA to move away from
its traditional decentralized management approach which placed most
development in field activities. Strengthening the program office and
helping it exert greater and more centralized control over the tasks it
delegates istunnatural in a highly matrixed-managed organization such
as the DLA, but we believe that it is necessary for the success of the
LSMP. While caution is advised regarding matrix management for the
LSMP, we also recognize the importance of obtaining input, cooperation,
and participation from the functional organizations. The program
office, which reports to the Office of the Director, was authorized to
a staffing level of 30 people. The program manager's responsibilities
included: the administrative and implementation plan; financial and
acquisition management; coordination and focal point with external
agencies; management of system architecture development and
installation; preparing status reports; contracting with outside
consultants; prioritizing work loads and scheduling; oversight of
training; and review of subprojects.
OCR for page 17
17
However, we cautioned that while the program management office
should be able to control development of systems, it would still take
strong direction, attention, and support from top management to enforce
the cross-functional cooperation and support needed to make LSMP work.
We also cautioned the DLA on the weaknesses of its matrix management
for such a program on the ground that its resources were geographically
end 'organizationally spread out and lacking in accountability. We
suggested assigning functional and technical people to the program
office for extended periods as a way to address this problem. We were
also concerned that the program office might become an island, isolated
from the rest of the organization, without the resources to do the job
by itself, and lacking the assurance that other organizations would
enthusiastically support the program and respond to its direction.
Strong Central Management
This program is of great consequence and promise. However, it is
sorely in need of leadership, authority, and an organizational focus to
move ahead with this worthwhile work.
We were recently advised that the Director, Defense Logistics
Agency, has directed several significant changes in his memorandum,
"Automated Data Processing (ADP) Modernization," dated 20 January
1989. In particular, the program office for the LSMP was merged into
the Office of Telecommunications and Information Systems (OTIS). We
believe that this move serves to consolidate what were previously two
different organizations with some overlapping interests into a single
entity that can be more readily held accountable for results. We
applaud this change and hope that it is the first of several
organizational changes that are needed to focus the DLA's automation
resources and its embrace of modernization.
The formation of the DLA from various military service organizations
has resulted in a great deal of autonomy within the various logistic
authorities. As a result, the geographically dispersed and autonomous
activities within the DLA tend to have differing goals and objectives
that optimize their own operations, but not necessarily those of the
organization as a whole. This has manifested itself through the growth
of multiple design activities and program managers located throughout
the agency and working independently on local requirements, even though
their functional operations may be similar to others. Central design
activities with responsibility for one or more automated information
systems are located at DLA facilities in Columbus, Ohio; Ogden, Utah;
Battle Creek, Michigan; Memphis, Tennessee; and Philadelphia,
Pennsylvania. Other design activities include the Defense Logistics
Services Center, Defense Technical Information Center, Defense Fuel
Supply Center, Defense Administrative Support Center, DLA Automatic
Addressing Systems Office, and Defense General Supply Center. As a
result there are numerous upgrades and modernizations taking place
concurrently with no central design, oversight, or control of these
activities. We recognize that each one is needed when viewed in
OCR for page 18
18
isolation, but we are concerned that if this continues the DLA will
severely limit what, if anything, it can achieve through the LSMP in an
integrated sense.
In order to ensure successful modernization, it will be
appropriate and desirable to consolidate authority over the
central design activities and those programs that have received
approval of mission need from the Major Automated Information
Systems Review Council, and to establish reporting channels for
them to that authority.
Responsibility, authority and accountability for program managers
and their organizational alignment should be decided and put into
effect quickly to avoid loss of momentum. At a minimum, the
organizational elements responsible for modernization should have
sign-off responsibility on system upgrades and enhancements where these
involve choice of computer languages, data base management, and
protocols that will assure interoperability. With such precise
accountability established for program results, users and planners
should confine their efforts to concepts, capabilities, directions, and
requirements. Functional managers and users must participate and have
a vote in the design, development, testing, and deployment process.
The functional managers must maintain appropriate staff capability to
define requirements, track progress, participate in test demonstrations
and critique the flaws in the process. The application of independent
and objective test demonstrations are critical to this process. A
"test bed" environment could be useful for users to conduct "what if"
experiments.
Information Resources Management
Industry regards the position of Chief Information Officer (CIO) as
much more than just an ADP manager. CIOs have to know their company's
business well enough to affect strategic decisions and to promote the
design of systems that support that business. Both technical expertise
and business knowledge are essential to design a technical architecture
that will fit the organizational architecture. Increasingly, the trend
is toward greater integration of systems, technologies, and data as a
means to improve efficiency and effectiveness. In the federal
government, this recognition and responsibility rests with the agency's
senior Information Resources Management (IRM) official (U.S. General
Services Administration, 1987a,b). In both the government and private
sectors, the challenges are similar and include coping with huge
investments in old technology; driving change from a position of
limited responsibility; planning developments that will protect future'
investments; finding ways to improve development productivity; and
developing their organizations. Even though such challenges are
formidable, the "reality" of urgent but short-term business needs
usually take precedence and consume most of the available resources.
OCR for page 19
19
Because the information relevant to each aspect of the
operations of the Defense Logistics Agency has significant
bearing upon the information relevant to every other aspect, the
agency needs to expand its view of information systems beyond
traditional automatic data processing and telecommunications
functions by considering information as a resource and creating
the organizational structures that will manage this resource.
In our review of the DLA, we found this distinction between ADP and
IRM lacking. Therefore, we strongly urge the agency to expand the
purview of the OTIS beyond its traditional ADP and telecommunications
(ADP/T) activities. The Paperwork Reduction Act of 1980 (P.L. 96-511),
and the Paperwork Reduction Re-authorization Act of 1986 (P.L. 99-500)
created and defined the concept of Information Resources Management
~ IRM) . Public Law 96-511 defined information as a resource and
directed agencies to designate a senior official for IRM. Public Law
99-500 added telecommunications to the definition of data processing
equipment. The Department of Defense appointed a senior official for
IRM and so did each of the military services. The Assistant Director,
OTIS, was appointed as the DLA' s IRM official in 1983. However, since
IRM responsibilities do not reside in a single organization, the
authority and accountability of the IRM official at the DLA is limited.
The DLA needs an IRM organization that focuses on the agency's
information requirements and understands the needs of its users as well
as the technological tools that can be applied. The OTIS can and
should develop into the DLA' s IRM organization and the agency should
recognize that this implies more than just ADP/T responsibility. It
requires a blending of applications knowledge and technology choices
from an information perspective.
DEFINING MODERNIZATION
When we began our review of the LSMP, we noted that even key
personnel in the agency did not have a consistent understanding of what
this program was to achieve and in what time frame. To our knowledge
only a small percentage of the agency's employees even knew that the
program existed. We urged the agency to set forth its goals and
objectives for the LSMP in a straightforward manner and to promulgate
this information throughout the agency. In September 1988, the prior
director released his strategic vision for logistics modernization
(Defense Logistics Agency, 1988a) that served to announce the LSMP and
what its implications were for the future of the DLA. We reviewed this
document and felt that it served to communicate a consistent top-down
view of the need for modernization, the broad changes that this would
entail, and the benefits that would accrue to both the agency's
employees and customers. Also in September 1988 the DLA released its
strategic plan, "Supporting the Armed Forces" (Defense Logistics
Agency, 1988b). We had urged and expected that this document would
provide specific and direct guidance to key personnel within the agency
regarding the plans for modernization. It neither attempted nor
accomplished this result. In fact, it listed the LSMP as one of five
OCR for page 20
20
modernization programs taking place within the agency. We also noted
that two of these modernization programs had been previously identified
as critical baseline enhancements, or short-term upgrades that were
needed to maintain effective operations. We believe that this served
to confuse the mission for LSMP both within and outside the agency, and
obscured the accounting of costs and benefits among the various
programs.
It was not until the release of the conceptual functional
requirements (CFR) in May 1988 that the LSMP began to be defined in
terms of what it might accomplish. In the meantime, review schedules
for the Major Automated Information Review Council (MAISRC) were
slipped as various upgrades, modernizations, and priority developments
continued outside of the LSMP umbrella. The agency went about business
as usual in upgrading and modernizing its systems as the vision of LSMP
grew more distant and its mission and goals became diluted by emphasis
on nearer-term programs. Our midterm report warned emphatically
against allowing this to happen, and suggested that close management
attention was warranted to avoid losing sight of the longer-term
modernization. By the Fall 1988, the LSMP languished and its program
office had become what we feared -- an island.
Refocusing
In November 1988, a new director assumed command of the agency. At
his urging, a cross-sectional group of senior DLA executives and
representative from the Office of the Assistant Secretary of Defense,
Systems, met off-site to resolve their differences and agree upon a
course of action that all would support. There were seven specific -
recommendations that the off-site management group chose to make to the
director of the DLA. They are listed as follows with our comments:
I. The Assistant Director, OTIS becomes the DLA's ADP czar.
We believe that the term ADP is obsolete and too restrictive.
We favor a senior IRM official for the DLA.
2. Merge the LSMP program office into the OTIS.
We agree with this recommendation to the extent that it assures
the LSMP will not be forgotten; that the program office will
cease being an island; and that the OTIS will acquire
accountability.
3. Central Design Activities (CDA) subordinate to the DLA's Systems
Automation Center (DSAC) which in turn reports to OTIS.
We agree with this recommendation because it establishes central
management.
OCR for page 21
21
4. Revise the reporting lines of program managers.
Managers of significant automation programs that have received
approval of mission need (MAISRC, Milestone l) should be
accountable to a central organization for the technical aspects
of the program. The OTIS appears to be most appropriately
positioned for this role and as a minimum should set DLA-wide
design guidelines.
Fully adopt the conceptual functional requirements (i.e.,
institutionalize them).
We agree with this recommendation.
6. Redefine the LSMP as a series of separate ADP projects as
opposed to a single agency-wide program.
We agree, but with the understanding that the program will be
executed as a series of separate projects coordinated under a
single broadly defined plan. One of the five major
recommendations made in our midterm report was that the LSMP -
best approached by looking at, and designing for, the whole
agency and then decomposing the program into smaller
sub-projects that can be implemented separately.
7.
Establish an internal MAISRC.
This appears to be a good idea for several reasons. It should
better prepare the agency for the DOD MAISRC process. It should
improve internal communications and coordination. It should
provide agency-level visibility and approval of agency-wide
automation initiatives.
These recommendations have led to what the DLA has called a
refocusing of the LSMP. By refocusing, the agency has recognized the
futility of trying to plan and cost-justify in advance costs for
equipment and services that are to be purchased significantly far in
the future so that details can not be known with any reasonable degree
of certainty at this time. Even so, a good deal of the motivation for
refocusing is thought to be attributable to budget constraints and the
anticipated unpopularity of large programs with substantial life cycle
costs. The motives for refocusing and the logic behind it appear to be
well founded. However, we are concerned that the elements are in place
for the agency to fall back into its operating mode prior to the LSMP
where systems were upgraded on a piecemeal basis with no real
improvement in the overall process.
The Missing Element
In our opinion, there is a set of steps that organizations typically
follow as they move toward the implementation of a long-term vision.
OCR for page 22
22
These steps are accepted in industry and also fit the planning and
program management structure of the Office of the Secretary of Defense
(OSD). In developing the LSMP, the DLA has acknowledged each of the
steps and has devoted effort at completing each one. The steps are as
follows, with reference to their respective DLA action:
A statement of the vision -- The Logistic Systems Modernization
Program (LSMP).
An analysis of the current manner in which business is
conducted -- The Business Area Analysis (BAA).
A model of how the enterprise will operate in the future,
organized into time-related phases from the near term through
the vision -- The Enterprise Model.
The changes necessary to implement the model -- The Business
Area Requirements (BARB) and the Conceptual Functional
Requirements (CFRs).
· What technology is needed and how will it be used -- The
information processing strategy presented in briefings to the
committee in November 1988.
· A strategy for managing the process of transition from the
present to the vision -- The Strategic Plan.
· Organizational arrangements and the assignment of roles and
responsibilities for carrying out the strategy -- Various
management directives. -
· Adherence to organizational oversight processes -- Recognition
of the OSD requirements for program oversight through the MAISRC
process and the formal Program Objective Management (POM)
process.
As noted above, the DLA has expended effort and resources in
addressing each of the steps. However, we do not believe that this has
been done with coherence of purpose or with the degree of senior
management involvement that is needed to cause organizational
components to pull together to make the effort productive and give it a
high probability of success. It appears as though the planning staff
diligently went about its tasks, checking off each of the steps but
doing them in a seemingly independent fashion. This resulted in a lack
of cohesiveness among the products produced. In our midterm report we
acknowledged that the DLA had done a great deal of detailed bottom-up
data gathering, but recommended more top-down analysis and point of
view.
OCR for page 23
23
The steps listed above involve three primary elements: information
and data, organizations, and technology. We have observed a
conspicuous lack of technology content in the-planning of the LSMP.
This may have been the result of an overreaction by the agency to the
Milestone 1 disapproval by the MAISRC, which found the proposed concept
too hardware-oriented and tasked the agency to develop a more
functionally-based concept. It also appears as though the DLA has
received conflicting guidance from the OSD: first, to emphasize
long-term functionality for one OSD group; and then afterwards to
emphasize details for another. In any case, technology elements took a
back seat when the DLA undertook a major effort to analyze and model
its business. We supported the business analysis effort in our midterm
report and we still believe in the value of such a process as long as
its results are put to use. The LSMP began as a technology-based
program, and it remains one. We urge the agency to see that the
technology elements of the program are properly included in its plans
by setting forth the technical architecture and agency-wide standards
that will translate the business planning into useful implementation.
Moving Forward
The MAISRC approved Milestone O for the LSMP in November 1986 but
withheld approval of Milestone 1 at that time. The DLA initially
expected to return to the MAISRC in March 1988 but slipped that and
several subsequently scheduled review dates. As of our most recent
briefing in February 1989, the DLA does not yet have a firm date set
for MAISRC 1 review.
During the period of our review, two smaller modernizations for the
logistics services center and the automatic addressing systems office
have been carried forward and received approval from the MAISRC for
Milestone 1 and O. respectively.
In order to avoid Jeopardizing the program and reemphasize
commitment to it, the agency needs to move forward with the
Major Automated Information Systems Review Council for the
Logistics Systems Modernization Program (LSMP) by scheduling and
holding to a date for review and approval of Milestone 1,
concept development for the first project under the LSMP -
~mbrella plan.
In presenting projects for MAISRC approval, the DLA should not try
to present plans and details that it cannot know at this stage.
Conversely, the DOD should not require the DLA to present details that
can reasonably be expected to be known only in future phases of the
MAISRC process. Refer to Appendix D for a description of-the six
phases of the MAISRC process. Instead of a single integrated program,
we suggest that the DLA maintain an umbrella plan and seek MAISRC
approval of subprograms to accomplish specific pieces of the CFRs in an
evolutionary manner.
OCR for page 24
24
Becaude the major benefits of modernization arise out of
improvements in process, the agency's modernization should be
guided by a unified concept rather than a collection of
individual and uncoordinated fix-it projects, performed on an
"as needed" basis, that retain the old processes.
By maintaining the overall mission of the LSMP, the DLA will have
based its needs on functional improvements, rather than individual
fix-it programs, and received some measure of assurance that the
program, and what it can achieve, will survive yearly~budget scrutiny
For this reason we also suggest that the LSMP be placed on the DOD's
formal Program Objective Management (POM) process because of its size
and multi-year execution period. -
Where to Start
The DLA should concentrate its modernization investment where it is
most needed and where it will produce the greatest benefits. We
suggest an approach where the agency selects and defines which of its
major business areas it will work on first. The CFRs need to be
converted into bounded and realistic projects that can be successfully
executed. Modernization subprograms need to consider and complement
other major DOD programs such as the Engineering Data Management
Information and Control System (EDMICS) which is part of the
Computer-Aided Acquisition and Logistics Support (CALS) initiative. In
this regard, the agency must make sure that its projects are achievable
in the "real world." Some of the CFRs can not be implemented presently
because they depend on policy decisions that are still to be made in
areas such as item visibility.
Because it is the largest part of the agency's mission and
because its systems are most in need of modernization, we
recommend that the materiel management and supply functions have
priority.
These functions are currently supported by the Standard Automation
Materiel Management System (SAMMS). This system supports the inventory
management functions at the agency's supply centers. It allows some
real-time interface with other on-line systems, but primarily uses
batch processing modes. SAMMS functions include distribution,
requirements, supply control, financial management, accounting and
billing, procurement and production, and technical cataloging. A
modernized SAMMS would improve materiel visibility, reduce inventory
and transportation costs, while increasing availability of supplied
items. SAMMS modernization, therefore, is one of the first major
subprograms that should come-forward under the LSMP umbrella plan.
However, there are three major components that need to be developed
under the LSMP: data bases, applications, and a technical facility.
Of these three, the data base design, and the agency's transition to
it, is by far the most significant undertaking, and relates directly to
how
OCR for page 25
25
the agency plans to conduct business in the future. SAMMS, like other
applications, can and should be developed and operated in conjunction
with an agency-wide standard data base and technical facility. We
suggest, therefore, that the design and development of data base
systems and acquisition of technical facilities be conducted
concurrently with application development. We strongly recommend that
the DLA design, develop, and acquire common technical facilities for
use throughout the agency and data base architectures that will support
its internal and external information needs. Given such uniform
technical platforms, the agency can develop and improve its
applications over time with greater assurances of long-term
interoperability among systems. This will also improve the
predictability and accuracy of on-going maintenance and development
efforts.
RESOURCES FOR THE LSMP
Costs
Lifecycle costs for the LSMP are currently estimated at $2.8
billion. Keeping in mind that these are lifecycle and not system
costs, the cost projections for new hardware and software are not out
of line with what has gone toward the modernization of other federal
agency information systems, nor should they be viewed as out of line
for the modernization of an agency the size and complexity of the DLA.
The agency's past investment in information has been massive, and it
currently averages an information management budget of about $300
million a year. Yet, underlying some internal DLA concerns with the
LSMP is the thought that a $2.8 billion modernization plan is simply
too large to take forward to OSD for approval. We believe that life
cycle costs should not be the only factor considered in the construct
and packaging of the LSMP, nor even the most important factor. The DLA
needs to consider the longer-term effects of technology-based changes
rather than evaluate them solely on cost savings. The agency's future
operations will depend upon a sound systems architecture being in place
to provide the vital infrastructure to support its mission.
The LSMP's life cycle costs should also be viewed in the context of
what the agency is planning to spend outside of the LSMP each year on
hardware and software upgrades and modernization of its ADP and
telecommunications (ADP/T) systems. For example, in Fiscal Year 1989
the DLA plans to spend about $87 million on hardware and software
acquisitions for ADP/T. This amount does not include personnel costs
or budgets for the LSMP. For the LSMP, the agency has budgeted only
$8.9 million for Fiscal Year 1989, or 9.2 percent of the total.
The DLA is urged to maintain its focus on the umbrella plan,
estimate the costs for each subprogram, and present these to the OSD
for MAISRC approval of Milestone 1. Concurrently, it must generate the
funds needed to support both the early development work and the
time-phased implementation of the discrete elements of the program.
OCR for page 26
26
This level of budgeting and planning is necessary to assure that the
MAISRC can give this program serious consideration based on its merits
as well as the likely consequences of failure to act on it.
The present level of LSMP funding for the next few years will do
little more than allow for continued study and planning -- a strategy
that will consume precious time, yield minimal results, and might even
jeopardize the DLA's business and information planning investment. In
Fiscal Year 1989, the LSMP has been budgeted for less than half of the
amount projected. For Fiscal Year 1990, the current budget
authorization is only 29 percent of the costs projected. Presently,
for Fiscal Year 1991 and beyond, the budget is unidentified. During
the five year period including and following Fiscal Year 1992, the
total annual budget projected for the LSMP is approximately $90
million.
As part of its funding strategy, the DLA should seriously consider
extending the LSMP's budget years and enter it into the Program
Objective Memorandum (POM) process. This will impose discipline on the
costing process. The POM process would require the DLA to acknowledge
its full costs to modernize while providing a vehicle for the agency to
gain the OSD's long-range support. We find it difficult to understand
how the LSMP can be taken as a serious program by the MAISRC without
adequate budget and POM support. We believe that the some holds true
for Congress, especially when the program is included in the budget
submitted by the President for Congressional approval.
The Role of Contractors
The DLA's technical staff consists of a significant number of
designers and programmers skilled on current systems. However, we are
not confident that sufficient skills and depth currently exist in the
DLA to successfully oversee and direct the development and acquisition
of the new systems, technical facilities, and data base architectures
that will be required as part of the LSMP. In our midterm report we
suggested that the DLA would need to use contractors as a way to obtain
the necessary expertise. We continue to hold this view.
However, we do not wish to see the agency become overly dependent
upon contractors. In our view, contractors can bring new ideas and a
knowledge of current technologies to the agency but they cannot be
expected to know the DLA's business or its strategic directions.
Contractors will allow for a transfer of technology to the agency's
personnel and should serve to augment rather than replace skills. We
favor decreasing the DLA's reliance on outside technical knowledge over
time and leveraging its contractor relationships into something more
like partnerships than skill supplements. This implies the need to
develop DLA's own resources, which can be best accomplished by
identifying key people; investing in their training; recognizing and
rewarding their achievements; challenging their skills; providing the
best tools; and by cycling staff between business and technical
assignments. Such a strategy will allow the DLA to retain an in-house
OCR for page 27
27
competence in information resources management rather than seeing such
competence depart when the contractor departs.
For the modernization program, the agency's acquisition strategy
calls for integration, design, and management by the DLA with -
contractor support. We agree with this strategy, but with the
understanding that the integration support contractor will be separate
from the contractors who work on the discrete modernization projects.
The DLA needs healthy doses of outside expertise in its integration,
design, and management roles as well as contractor support in
day-to-day systems maintenance activities.
ISSUES IN THE DEFENSE LOGISTICS ENVIRONMENT
Our midterm report identified the need for certain external policy
guidance from the Office of the Secretary of Defense (OSD) regarding
problems that the DLA could not resolve on its own. Chapter 3 of the
midterm report, "The Defense Logistics Environment," dealt with this
issue in detail. At this time, the major issues are still materiel
visibility, excess inventories, and stocking policies across all the
Services and the DLA. Admittedly, these are difficult and complex
issues, but to the best of our knowledge, they have not yet been
addressed by the OSD and no guidance has been provided. In this final
report, we wish to draw attention to other issues external to the DLA's
control that also warrant attention.
Open Competition Versus Standards
This issue involves the effect that open competition has on the
ability of an agency to plan for modernization, and on the plans
themselves. It applies not only to the DLA but to most federal
agencies. In general, open competition is a valid and effective means
of improving the value received from a procurement. However, we
believe that when practiced in full and free manner it also constrains
an agency's ability to develop long-range plans by introducing unknowns
into the acquisition process. Furthermore, we believe existing
regulations that make it quite simple to delay an acquisition through
the use of protest action further exacerbates this situation.
DLA is committed to open competition in acquiring new information
systems. However, our advice to set standards and guidelines, and the
DLA's efforts to do so, are often at odds with permitting any supplier
to marginally bid what can be regarded as a functional equivalent.
Accepting the lowest bid can easily lead over time to system
acquisitions that are neither interchangeable nor interoperable with
others. In our view, maintenance and support burden increases with the
number of different system designs in operation.
Open competition is certainly desirable for commodity procurements
but a more disciplined application may be needed for the complex
information technologies that must be acquired over a long period of
OCR for page 28
28
time and applied to serve the unique needs of an organization. Private
industry has tong recognized that standardizing on certain hardware and
software was necessary to prolong the life of its systems investments;
assure that information could be shared and moved to where it was
needed; and hold down long-term maintenance and development costs.
When and where possible, industry seeks to acquire components in a
competitive environment but generally controls and specifies its
requirements in a way that effectively limits responsive bids to those
firms that den offer and perform what is required. Typical practice in
receiving proposals is not to further evaluate bid offers if they would
jeopardize existing investments or complicate operations by diverging
from organization-wide standards. However, in the government sector,
agency managers faced with multiple acquisitions of information
technology have limited motivation and resolve to select a technical
solution and stick with it.
The DLA must be aware of these factors in choosing its acquisition
strategy. A single integration contractor for the LSMP would serve to
protect the agency from the uncertainties of multiple procurement
actions but would also relinquish some control of the program. By
awarding multiple competitive contracts, the DLA will increase funding
flexibility and may satisfy critical needs sooner, but it must be
resolute in ensuring the acquisition of a common technical facility for
its various applications over a long time. In order to accomplish
this, the DLA's contracting officers must be qualified to award and
administer technology-based contracts and its procurement
specifications must conform to DLA-wide standards.
Acquisition Protests
Also of concern to us are the regulations that almost seem to
encourage unsuccessful bidders to protest a procurement award
decision. By establishing and maintaining good communications with its
bidders, the agency can help to avert protests that can arise from
ignorance or misunderstanding. If protests occur, they should be
settled within the BLA rather than escalating beyond the agency. This
places a premium on the skills of the contracting officer and his or
her technical representative, especially their technical competence and
management maturity. Because so many of the DLA's procurements are
routine and commodity-based, contracting officers are not presently
required to possess senior management qualities and technical
competence. We urge the agency to recognize this need so that its
modernization acquisitions will be handled by personnel who have the
knowledge, responsibility, accountability, and authority to do so. We
also urge the director and the OSD to speak out on counterproductive
mechanisms in the procurement process, and to initiate measures that
provide ample safeguards while allowing progressive, well-planned
modernization.
OCR for page 29
29
Turnover at the Top
The LSMP as presented to us involves a major long-term commitment by
the DLA to modernize the way it carries out its mission. This vision
needs to evolve and develop into implementing programs over time and
through changes in top management. In this effort, we underscore the
need for vision, perseverance, and flexibility in order to realize the
full benefits that modern information systems and technology will bring
to the agency now and in the years to come. The vision and work of
modernization will necessarily span several changes in the agency's top
management positions that are held by military officers. The
modernization process must also expect that priority initiatives will
need to be addressed as they arise from time to time. The agency must
continue to perform its functions without disruption during the time it
modernizes its systems and processes. Continuity, therefore, is a
consideration that must be addressed in the conduct of this program.
In addition to creating permanent positions and institutionalizing the
process, the agency may also wish to consider setting up an external
advisory committee, a mechanism similar to the Defense Science Board,
to maintain continuity and momentum.
Exporting Accomplishments
During our review of the DLA's systems, we noted that the agency has
accumulated considerable expertise in the systems it had developed. In
addition to their development effort, the agency has continually
updated and improved these systems. We feel that the agency is taking
steps to manage the development and changes it is making to its
hardware and software systems through documentation, code and release
version control. These efforts are incorporated in a system
engineering discipline known as configuration management. We stress
the importance of configuration management in allowing the agency to
operate its geographically dispersed data processing and design
activities, and to coordinate the software and hardware used in common
by the DLA's major business areas. Configuration management is a
process that occurs through the systematic identification, control and
audit of system characteristics. It follows the production, review,
and acceptance or rejection of a product. Configuration control
requires the evaluation, coordination and approval of proposed changes
to an established baseline of software items. By effectively employing
configuration management, the DLA will be in a position to confidently
export its best systems to other DOD entities that have a similar need
but less expertise. We understand that the DLA's contract
administration and personnel systems can be used by other DOD
entities. We support and encourage the common use and sharing of
systems. In this way, the DOD can propagate its best systems and avoid
needless duplication.
OCR for page 30
30
REFERENCES
Defense Logistics Agency. 1988a. Strategic Vision. Logistics
Modernization. A Message from the Director. Alexandria,
Virginia: Defense Logistics Agency, Cameron Station.
Defense Logistics Agency. 1988b. Supporting The Armed Forces.
1988 Strategic Plan. Alexandria, Virginia: Defense Logistics
Agency, Cameron Station.
U.S. General Services Administration Information Resources Management
Service. 1988. Washington, D.C. An Evaluation of the GRAND
DESIGN Approach to Developing Computer Based Application Systems.
U.S. General Services Administration, Information Resources Management
Service. 1987a. Washington, D.C. The Senor Federal IRE Manager:
Major Roles and Responsibilities as we move into the 1990's.
U.S. General Services Administration, Information Resources Management
Service. 1987b. Welcome to Information Resources Management in
the Federal Government . Washington, D.C.
Representative terms from entire chapter:
midterm report
