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 13
III. DISCUSSION OF FINDINGS
Finding #] - Air Force Progress on Base-Leve] Automation.
Within the limits of its current resources, strategy, and approach
the Air Force is doing a good job of base-level automation' but much more
remains to be done.
sound.
a. The concept of a standard base-level automated support system is
Several decades ago, the Air Force made a fundamental decision on
base-level automation. Tt decided there was a fundamental advantage in
providing a standard data processing facility to each Air Force base, and
that most common support systems would operate using that facility. This
decision was reaffirmed and strengthened when the Air Force decided to
enter into the Phase IV program to reequip the standard base facility while
consolidating a number of different functional systems into a single ADP
facility. The Committee agrees with the decisions made by the Air Force
over 20 years ago:
0 To establish the air base as the point of focus for the build-
ing of automated support systems,
o To maintain a set of standard configurations for ADP hardware
and software at all bases.
In the Air Force, as in most organizations, the first introduction
of automated data processing was to support the accounting and financial
functions. Computers installed for this Purpose then began to support
other functions. As the capacity of these was exceeded, equipment and
software were acquired by other functional components of the organization
to support their own activities. Starting in the early 1960s, the Air
Force began to standardize its base-level automation support, first with a
mainframe system to support supply, then with a separate system to support
all other standard functions.
At the time the planning was undertaken for Phase IV, there were
four major functions within the Air Force that depended upon automated
-13-
OCR for page 14
support: accounting and finance, personnel, supply, and maintenance. Each
of these is an Air Force-wide function, and the responsibility for general
guidance and policy resides in its own separate Headquarters-leve] organi-
zation. Each, however, directly supports and affects the activities of the
air bases and the responsibilities that reside there. Hence, the Air Force
elected to make the individual air base the focus of the automated data
processing activities that support the corporate functions at that base.
In conjunction with Phase IV, the Air Force not only provided for
modernization and capital improvements, it also took other significant
steps. For example, the Air Force:
o Preserved the concept of the base-level system as the primary
unit around which to organize its automated support activi-
ties.
O Provided a single facility -- hardware and system software --
to support the four "charter" functions of accounting and
finance, personnel, supply, and maintenance, and to support
the many other functions at a base that were in need of auto-
mated data processing.
O Established the concept of a standard family of computer
equipment to be provided at all bases, having flexibility for
variations among bases in size and mission.
The Committee considers that these decisions were valid. Their
effect has been to put the Air Force in position from which to move toward
further modernization and functional improvements.
b. The Phase IV capital replacement program is going well, but
additional, we]~-directed efforts are needed to bring it to an orderly
conclusion.
The original strategy for Phase IV was:
o To rigidly control the software functional baseline of the
standard base-level applications as of 1980;
o To replace obsolete UNIVAC and Burroughs hardware with the new
Phase IV environment;
0 To convert the software functional baseline to run within the
Phase IV environment and apply limited updates as necessary;
o To do all this without loss of functional capability or
degradation of response time.
These objectives have been or soon will be realized at all instal-
iations involved in Phase IV.
-14-
OCR for page 15
As with any new installation of this magnitude, there are "bugs"
to be ironed out and some "fixes" are needed. In addition, there are
issues that were not fully addressed during the difficult planning and
implementation of Phase IV that must now be faced -- for example, the
training of personnel for the new systems. There also are many simple
improvements that can be made to systems now in place, that will lead to
greater efficiency and broader functional capabilities.
The Air Force is aware of the "bugs" and the "fixes" needed and of
many functional improvements that need attention. A number of short-term
efforts have been initiated to help users get the maximum benefit from the
current Phase IV configuration or simple variations thereof. However, the
Committee observes that no single Air Force document now exists that des-
cribes an integrated plan, with priorities, schedules, and resources
identified to accomplish the fixes and improvements.
The Committee's suggestions on development of a Phase IV post-
implementation enhancement plan are provided in Recommendation #l, Section
IV of this report.
c. The Air Force has taken important initial steps to evolve toward
a more powerful post-Phase IV base-level system' but significant tasks
remai n.
The Air Force has done an excellent and imaginative job of using
the so-called "requirements contracts" to augment the Phase IV configura-
tion and to provide distributed processing power and direct support of
users at the bases. Some of the functional area applications, such as the
Core Automated Maintenance System (CAMS), are being incrementally improved
beyond the capability achieved by merely converting the old software to
Phase IV. A continuation and broadening of such efforts would exploit the
Phase IV hardware/software foundation and provide significantly higher
levels of base automation. Even though the Air Force was directed to buy,
rather than lease, the Phase IV equipment, it appears to the Committee that
there are possibilities for using the flexibility of the Phase IV con-
tract's "technical replenishment clause" to update and expand the Phase IV
portion of the standard base configuration, and the Air Force is
encouraged to use this flexibility.
The Air Force's efforts are almost certainly constrained by the
fact that it does not have an announced policy for base-level automation
spelled out in a declaratory document. Such a document is needed to enable
internal Air Force organizations to plan, program, justify, fund, and
coordinate their activities. In addition to looking into the policy area,
the Committee thought it important to examine the Air Force's strategy for
base-1eve] information systems. Al though there is no single strategy
document, based on the documents, briefings, and programs that have been
provided by the Air Force it appears to the Committee that the strategy
has been:
-15-
OCR for page 16
o To supplement the Phase IV environment with a range of micro-
computers and minicomputers, and LANs to convert them into a
distributed processing network.
O To provide an added range of base-level facilities, such as
electronic mail, ad hoc on-1ine cross-fide inquiry, facilities
allocation, scheduling packages, etc.
O To provide an after-the-fact technical architecture to con-
ceptually tie together the varied processors, terminals, work
stations and data bases into a de facto network with distri-
buted processing qualities.
Having built this network and converted old functionalities,
the major unintegrated single-function support systems (such
as supply, maintenance, personnel, finance) are to be stream-
lined and upgraded incrementally, with interfaces to and auto-
mation of several previously manual functions. In some cases,
such as finance, the set of old systems will be augmented by
an entirely new system, namely the base-level component of
the Comptroller Office of the Future (GOOF). But in mainline
logistical systems, upgrades would be incremental.
This is inherently a suboptimal strategy. What is needed
is a standard base-level automation configuration for 1990, with a
number of different type of computational equipment allowed. The
instead
limited
Committee
.
noted that the Air Force has begun to address this need through a well
conceived effort within the Air Staff (HQ USAF/ST) to develop a standard
base ADP hardware, systems software, and communications architectures We
applaud this start and would like to see it go further and faster.
date, however, the efforts have been mostly contemplated and do not
the needed roadmap to commit the development of the future teas
information system. The architecture should be carried to a
functional detail, showing where the individual syste'
electronic mail switching; common storage; data entry ~ ~
To
provide
e-level
Eve 1 no
it of the future base-level
Id be carried to a level of
Dual system functions (e.g.,
data entry) wild be accommodated
on a base, and at which processors or work stations. We would like to see
that architecture fleshed out and then translated into a standard, upgraded
base configuration. In so doing, we expect the Air Force will find that
the goal of a distributed environment, with large shared data bases, com-
prehensive protection, and decentralized processing, poses significant
technical problems. These must be resolved before the Air Force can pro-
ceed with implementation of the future base-1eve] architecture.
stributed environment, with
The development of a base-level architecture is not far enough
along for the Committee to judge its potential success. It is our judg-
ment, however, that considerably more attention must be given to the
overall systems aspect of the architecture. It is not sufficient to
provide communications networks and processors which can communicate over
those networks. Questions of how to decompose applications to run in a
distributed environment, how to share data bases and still protect the
data from unauthorized access, and how to utilize a heterogeneous array of
-16-
OCR for page 17
processors in an integrated way must be dealt with.
As described above, Phase IV was essentially intended to replace
functions performed by older systems with the same functions on newer
common hardware. Only limited capacity was added for growth and no capac-
ity was added for reserve against such surges as might be triggered by
crises or preparations for war. Only limited capacity was added for the
improved capabilities that might serve more users. There were many good
reasons at the time for such caution in the planning of Phase IV, but the
effects of the decisions that stemmed from that caution are already being
felt. For example, the Air Force soon may find itself with insufficient
computing capacity at its bases. In addition, the Air Force is now dis-
playing strong centrifugal tendencies, as exemplified by the variety of
applications and equipment that have been identified and in some cases
procured.
The Committee has heard anecdotes and unverified quantitative
arguments suggesting that at some bases the workload on the present Phase
IV installations already exceeds capacity. The Committee has not enough
data to judge whether this overload could be relieved by the use of more
modern and efficient software. The Committee has no doubt, however, that
the presently installed data processing capacities will inevitably be
exceeded as a result of operational demands for increased capacity,
improved quality, and better functional capability.
Briefings by Sperry Corporation representatives indicated that a
50-fo1d increase in the capacity of a Phase IV system might be achieved by
aggressively exploiting the "technical replenishment clause" of the Phase
IV contract. But it must be noted that an annual work load growth rate of
only 22-30 percent can easily consume a capacity growth factor of 50 in
less than the permissible 20-year duration of the Phase IV contract.
The Committee is more and more convinced that the Air Force must
re-examine base-level projections and assure itself that the growth in the
Sperry family of equipment, the evolving of tease-1 eye] architecture, and
the attainment of a cohesive on-base automation environment (recommended
elsewhere in this report) can accommodate the expected high growth rate
through at least the year 2000. Otherwise, the Air Force will find itself
facing another major equipment upgrade, with the inevitable disruption of
on-base activities and the enormous effort required at all levels to
specify, advocate, and install such an upgrade. We cannot predict tote]
capacity required, except to note that it is much more than current capac-
ity, it will change radically in time, and it must be adequate to run
MACCOM-specific applications as well as the standard system.
There are a number of technical arguments to support standardi-
zation and configuration control, but the fundamental point is a military
one. As time goes on, both wing and base functions become more and more
dependent on automation support, to the point that the wings/bases cannot
operate without base-level ADP. As stated earlier, the Committee agrees
with the original decision to provide a standard ADP environment to be used
OCR for page 18
in developing systems for individual Air Force bases. Furthermore, we
believe the arguments for establishing a standard configuration for the
base-level automation environment have become stronger, not weaker. Unless
the base-level automation system configuration is standardized, with strict
controls over the number of different hardware and software options that
can be employed, the future development and maintenance of hardware and
software will become a nightmare and hardware backup prohibitively
expensive.
The Air Force can procure the elements of the "standard configura-
tion" referred to above under current and planned contracts: under the
"technical replenishment clause" of the Phase IV contract for the improve-
ments that Sperry is offering or planning in the near future; under the
Air Force microcomputer requirements contract; under the planned mu]ti-
station suPermicrocomputer buy; and perhaps under the advanced mini-
computer purchase. Once the desired configuration is identified, it can
be compared to the current configuration and inadequacies resolved in a
consistent and logical manner.
To express our thinking metaphorically, every base must have at
least one standard automation "toolbox" and must use only the contents of
the toolbox for specific base functions. The toolbox should include the
right "tools" both for now and the future. Obviously, the toolbox wild
change with additions and improvements, but there must be a controlling
authority that monitors the standard base automation system configuration
and the contents of the toolbox to keep up with changing needs and new
technology.
The Committee's suggestions for establishing a standard base-
leve] automation configuration are provided in Recommendation #2, Section
IV of this report.
Finding #2 - A Window of Opportunity.
The Air Force has a window of opportunity to rethink and reorganize
several of the base-level functions for the 1990-1995 time period, and to
specify information systems to support these functions.
In many ways the base-level information systems resemble a set of
"stovepipe" systems, each supporting a single base-level function. Many
of the current base-level information systems, particularly those which
were converted from the old Burroughs or UNIVAC environments to operate on
Phase IV equipment, represent implementations of at least five to ten year
old functional requirements.
There is evidence, however, that the Air Force has taken steps to
solve the problems of (~) out of date functional capabilities and (2)
excessive functional compartmentalization flack of integration).
The Air Force had been upgrading both the functional specifi-
cations and the implementing code continuously up to the time the "base
_ ~ ~ _
OCR for page 19
line" was rigidly controlled during conversion to the Phase IV facili-
ties. This evolutionary upgrading continues for current systems, in some
cases incorporating major improvements, such as CAMS. For other func-
tional areas, such as personnel and finance, the Air Force reports it has
reexamined the functional requirements at the base, command, and Air Force
levels and has designed new systems (including base-level components) in
response to new requirements. The Committee has not yet investigated
these reports and associated new systems. For each functional area -- 27
in all, according to the HQ USAF Program Management Directive -- the Air
Force has undertaken to produce Base Information Analyses (BlA). Although
the various BlAs differ somewhat in detail, they appear to be intended to:
o Document current base-level information architecture in each
functional area.
O Identify problems in the current information flows.
O Recommend new system developments and/or enhancements to
solve these problems.
In regard to the lack of integration, the Air Force also is
traveling multiple paths in search of solutions. They are analyzing
required interfaces, such as that between maintenance and supply, and
identifyinq improvements that might be made to foster integration. But
the Air Force recognizes that some interfaces cannot be identified in
advance, such as those required to provide management information in
response to as yet unspecified and probably presently unknown queries by a
Wing Commander or an operations officer. Therefore, the Air Force is
developing an architecture that it is hoped will allow all work stations
to access the various computers on a base and communicate with each other
and support inquiries and report requests which extract and combine data
from more than one functional data base.
In short, the Air Force is attempting to build on the original
Phase IV concepts to Provide modernized base-level capability that would
(~) provide a standardized and highly integrated canonical base-level
environment (especially in hardware, systems software, and communications),
and (2) institutionalize a process that would permit evolutionary changes
in the development and integration of individual functional systems.
The Committee was impressed with the Air Force's efforts in
following the course outlined above, particularly in the base-level ADP
environment and the maintenance system. I t appears to us that these
efforts are weld launched and offer hope for successful implementation.
But we believe that the Air Force may miss several extraordinary opportu-
nities if present efforts are not broadened. These opportunities, which
exist in large part because of the current and imminent near term successes
of the base-level automation program, are:
_] 9_
OCR for page 20
to rethink and redefine base-level logistical requirements and
the supporting information systems for the mid-199Os; and
to analyze and define crisis and wartime needs for information
of wing commanders, their staffs, and squadron commanders in
carrying out their units' military missions.
There is a good example of the value of this approach in modern
commercial, highly automated manufacturing companies like General Motors
and Genera] Electric. At first, they concentrated on providing modern
computer-assisted design (CAD) technology to automate the design function,
computer-assisted manufacturing (CAM) technology for the manufacturing
function, and automated interfaces between the two. Such automation
produced significant benefits, just as independent automation of the Air
Force supply and maintenance functions has provided significant benefits.
But these companies now realize there are limits to the available
benefits from piece-wise automation; they realize they were in error apply
ing the most modern technology without re-thinking the underlying pro-
cesses they were automating. So, many firms are now reexamining the
entire process, without the arbitrary division into design and manu-
facturing. Presently, in advanced enterprises, the division is between
technology and production, with the former providing integrated design and
manufacturing productivity aids, and the latter being responsible for
applying those aids to what has become a "seamless" (functionally
integrated) process of design-manufacture. This new process, and the
supporting information systems, almost certainly would not have happened
by extending the old processes and building on the odd way of thinking in
an evolutionary manner.
a. Base-Leve] Logistics. The concept of process re-examination is
relatively easy to explain on the logistics side. A major improvement in
Air Force base-1eve] logistics is already in progress with CAMS.
Improvements in the supply information systems will allow base-level
maintenance personnel to enter requests for parts themselves, and find out
availability and expected delivery time. AFEC item managers will be able
to track status and location of their assets worldwide, in a timely
fashion, and make appropriate decisions.
In part because of the projected improvements in both base-level
and wholesale level information systems, the Air Force is able to consider
radical improvements in its maintenance, supply, and transportation strate-
gies. In maintenance, the Air Force is already contemplating the shift
from a strategy of repairing as much as possible on the flightline to one
of replacing as many components as possible on the f~ightline, while dis-
tributing and repairing the components at intermediate J eve] maintenance
centers.
The coming availability of more timely and accurate supply infor-
mation makes Possible major changes in supply philosophy as well. One such
change the Air Force is contemplating is a shift from a "demand-pull"
-20-
OCR for page 21
strategy of responding to requests from the bases, to a "supply-push"
strategy initiated and managed by the item manager and the MACCOM-level
supply officers. Similarly, major changes in intratheater transportation
are also in planning. One example is the European Distribution System,
which is analogous to the well-known Federal Express hub concept.
This plethora of new information processing technology and major
shifts in logistics strategy, suggests that a reorganization of the supply,
maintenance, and transportation functions into an integrated logistics
process may be in order and, more significantly, be possible. In any
event, the Committee believes there is sufficient probable cause to
consider afresh the information requirements for an integrated base-level
logistics system. We favor a top-down perspective, rather than a bottom-
up, incremental perspective of separate "stovepipe" maintenance, supply,
and transportation systems.
b. Base-Leve] Operations/Combat Support. The concept advanced here
is more difficult to define for base-level operations/combat support. How-
ever, the Committee believes opportunities exist for merging the wing and
squadron-level operations/combat support system with some of the base-
level functional systems, as the basis for a significantly enchanced wing/
squadron-level mission support system. One complication arises from the
fact that the operations/combat support systems will be different for each
MAdCOM. Another is that the missions they support in peacetime (such as
training and peacetime transportation of Personnel and material) are in
many cases quite different from the crisis and wartime missions. As a
result, the operations/combat support analysis will be more difficult than
the logistics analysis, it will take longer and it likely wild involve
several iterations of top-down guidance and bottom-up statements of need,
followed by top-down direction.
We have not had the opportunity to review the MACCOMs' operational
plans and the information systems designed to support them. Nevertheless,
several observations lead us to believe that mission support functions are
relatively neglected, and to suspect that a useful new high-1eve] system
could result from a new functional analysis of mission support requirement
on a MACCOM-by-MAdCOM basis.
_
o Little of the Air Force's current base-level documentation
and design effort appears to be devoted to operations, as
such, and of that, the greatest part is directed at peacetime
activities such as scheduling of personnel and aircraft for
training flights.
We have detected a pessimism on the part of Air Force officers
regarding the feasibility of predicting the wartime needs of
commanders. While we share that view when applied to high-
leve] commanders, we do not believe it would be exceedingly
difficult to predict the types of information that wing and
squadron commanders and their battle staffs would need to
support operations. This is especially true for the planning
-21-
OCR for page 22
of the second and later sorties of combat and combat support
aircraft in non-nuclear warfare situations.
O The top-down command and control and intelligence systems are
planned and designed in isolation from the maintenance and
supply systems, yet wing/squadron commanders will have to
combine information from both in order to support military
operations.
Evidence of the need for automation support for wing and base
commanders is provided in the effort introducing ADP through the Advanced
Concepts Base Program, which seeks to:
o validate the effectiveness and applicability of new technology
in an operational environment;
o demonstrate practical improvements in support to functional
users;
o evaluate new processes and strategies; and
o avoid wasteful duplication of effort associated with uncoordi-
nated prototyping.
This small effort has succeeded in marshalling operations and
combat support personnel to increase their effectiveness through the use
of ADP. The group works on a bottom-up basis and reports to the Base
Commander. This direct line to support the operations/combat support
mission should be encouraged to stimulate prototype applications for
evaluation by the information systems community for integration into the
standard base-level automation environment. Those who have operations/
combat support roles will specify what they need to do their jobs better.
In this way, the on-going needs of operations will be satisfied and such
efforts can become candidates for inclusion in the total base-level automa-
tion support system. The members of the Advanced Concepts Base Program
team facilitate this concept without being information systems officers.
The Committee has not seen much evidence that the Air Force
appreciates its "window of opportunity" to develop new, top down functional
designs for a base-level logistics system and a repertoire of command-
sPecific wing/squadron level mission support systems.
The Committee's suggestions regarding functional requirements
analyses for new base-level operations/combat support and logistical
systems are provided in Recommendation #3, Section IV.
Finding f3 - Dependence of Wing/Base-Leve] Units on Automation Support.
Air Force wing/base-level units are critically dependent on ADP
support to fight effectively, but the necessary actions have not yet been
taken to ensure these system capabilities are available for combat depioy-
ment and operations.
-22-
OCR for page 23
Most of the Air Force officers with whom we have worked will
acknowledge that this observation is true. That is, without portions of
the supply, maintenance, and operations systems and data, plus command and
control and intelligence systems, the Air Force cannot deploy or sustain
military operations. Yet, evidence abounds that this observation is not
reflected in current systems design, in equipment selection, and in
practice:
o Neither the deployable hardware nor software for the Combat
Support System (CSSJ "formerly Deployable Combat Supply System
(DOSS)] is used routinely in the peacetime base supply system.
When deployed, the CSS provides a lesser functionality than
that which is available in the peacetime system supported by
Phase IV. Additional effort and resources are required to
ensure that software is maintained up-to-date and supply
personnel are adequately trained on both the peacetime and
deployable systems. Due to these constraints, it is antici-
pated that supply personnel will experience great difficulty
in switching from the peacetime mainframe system to the
deployable microcomputer system in time of crisis.
O There is no standard deployable maintenance or operations
system.
o It is obvious that back-up repair and reconstitution of the
base-level ADP environment becomes increasingly more difficult
as the number of non-interchangeable essential components
increase. Additionally, in this case there are significant
problems of maintenance and cross-training. For these reasons
alone, the requirement for continued ADP operation in a war-
time environment should lead to a major reduction in the
number of different brands and configurations of components
of the base-level ADP environment. In fact, the opposite is
happening. Yet another and different minicomputer is being
procured for the deployable supply system, yet another unique
architecture and equipment is being contemplated for its base-
leve] personnel system, and the civil engineers are planning
on still another type of computer system.
The foregoing indicates that the Air Force is employing piecemeal
measures to support deployment. Instead, it should take deployment into
consideration as a fundamental objective of at J east its logistic and mis-
sion support systems, without which the Air Force cannot effectively fight.
The Committee's suggestions on automation support for combat
operations are incorporated into Recommendations #2 and #3, Section IV.
Finding #4 - Development/Maintenance of Base-Level Information Systems.
The Air Force can realize significant technological improvements (and
probably save money and manpower in the long run), if it changes the way
it develops and maintains base-level information systems.
-23-
OCR for page 24
The Committee notes that the Air Force has devoted great efforts
to the transfer, conversion, and transition of base-level systems from the
old UNIVAC and Burroughs computers to the Phase IV computer systems. In
the main, this effort focused on the preservation of the previous level of
functional capability.
From the earliest days of this study the Committee noted it had
not observed planned, consistent efforts to upgrade applications produc-
tivity. There was little evidence to indicate use (or planning for the
use) of modern software development tools such as fourth generation
languages, relational data bases, screen-driven program generators,
application Prototyping, and the like. New applications software
development and maintenance technology is available in the commercial
sector, but its wholesale transfer to the Air Force tease-1 eye] environment
is being hampered by a major capital (hardware) replacement program that
does not include modernization of the application software.
The Committee also concluded that little thought has been given
to the use of commercial contractors in this area, even though many have
modern tools that could expedite and improve software development and
maintenance. We believe each new application and modification should be
engineered so that it will be easy to use available productivity tools,
introduce user-friendly techniques, develop compatibility and interfaces
between existing and planned systems, reduce prototype and development
time, and improve user participation in the development process.
The Air Force personnel who spoke to this topic expressed a
recognition of the desirability for movement toward more modern
approaches, but expressed difficulty in making significant progress in
this area. The difficulties appear to stem from the following:
o The tradition and policy of using in-house personnel for
software applications development at Gunter AFS tData Systems
Design Office (DSDO)], Tinker AFB "Command and Control Systems
Office (CCSO)], Lowery AFB tAir Force Accounting and Finance
Center (AFAFC)], Randolph AFB tAir Force Military Personnel
Center (AFMPC)], and at the MAdCOMs (for base-level MAdCOM-
unique software), as well as heavy reliance on COBOL alone as
a development tool.
O The reported difficulty in justifying funding to acquire
software productivity fools, commercially-available software
packages, or to hire contractors. The chief difficulty
appears to be the policy requirement to forecast savings in
personnel, manpower spaces, money, or other areas as a pre-
requisite for approval of requests to acquire such tools and
capabilities.
O The fragmentation of base-level efforts as cited above, where
numerous functional entities, commands, and staff agencies are
-24-
OCR for page 25
planning independently or semi-independently to upgrade or
modify existing systems, replace existing systems, or acquire
totally new systems to be located at the base-level, in some
cases using computers that are not directly compatible with
Phase IV.*
In its deliberations on this finding, the Committee felt that the
Air Force would want to adopt the following objectives when considering an
appropriate course of action:
o To facilitate the introduction of new technology to support
base-level applications. This should include such technology
as software engineering productivity tools, modern applica-
tions systems architecture, and techniques for enhancing and
supporting near-identical software installations at over 100
sites.
o To increase the use of commercially available software
packages as major parts of the base-level systems.
o To drastically reduce development time and costs.
O To reduce the dependence of information systems in general--
and base systems in particular--on "blue-suiters" who are
already in short supply in the Air Force and who are likely
to be increasingly difficult to retain.
O To strengthen the management structure and process for base-
level automation, in order to ensure attainment of the
preceding objectives.
The Air Force faced a similar problem in the development of its
weapons systems soon after its creation. The increasingly specialized
needs of military aircraft and electronics made it difficult for the Air
Force to order off-the-shelf subsystems and integrate these directly into
weapons systems. In response, the Air Force shifted to a process in which
it spent more time and effort specifying complex weapon systems required to
support its missions. The Air Force established a System Program Office
(SPO) for each major weapon system and it delegated broad authority to the
SPO to ensure the successful completion of the program. The Air Force
also shifted more responsibility--once the weapon system had been specified
--to a single prime contractor, who not only would assemble but also would
* Examples are the Automated Personnel Data System (APDS-~), which will
use IBM Series ~ machines, and the CSS, which will use the Burroughs B26.
In neither case is it possible to interact with the Phase IV systems
without additional software, or to take advantage of the Phase IV data
base management system (DBMS).
-25-
OCR for page 26
specify and procure the necessary subsystems, including those available
off-the-shelf. And today, through a highly interactive process involving
the SPO, the users, the prime contractor, and others, the Air Force contin-
ues to specify enhancements even after initial procurement under which the
contractor updates the weapon systems through subsequent models.
As with weapon systems, today's modern information systems in the
base environment are complex in purpose and in technology, they are very
expensive, and they must support commanders and managers in both peacetime
and wartime situations. The Committee has had the opportunity to observe
the multiple activities that affect the present base-level systems; we
believe this is the norm. A further consideration is that various system
actions will be in different stages of system life and will require close
engineering, management, configuration control, performance review, and
budget controls to insure the Air Force receives acceptable results.
If the Air Force were to consider base automation as it does
weapon systems, we suggest it would decide such an approach requires the
designation of a SPO and the use of a prime contractor.
The Committee strongly believes the total base-level information
systems complex must be addressed as a single management entity. Only in
this way there can be proper planning, design, production, and coordi-
nation, as weld as the assurance that compatibility will be preserved
among the many functional systems supporting operational base activities
and all the other functional entities that reside and operate on a base.
The concept requires configuration control, standard maintenance
and operations procedures, and use of the latest technology for imp~e-
menting these procedures. For the Air Force base-level automation problem
in particular, centralized, automated support is required for over lOO
package installations, a situation faced by vendors of commercial pack-
ages on a daily basis. Considerations of cost, development time, and the
unavailability of Air Force information systems personnel, in our opinion,
call for much more extensive exploitation of commercial packages and
available application development aids. What is new is that such packages
are now available for elements of many base-level systems. The Committee
does not believe that piecewise introduction of new tools would achieve
success and suggest the Air Force may have to introduce them in a broader
way, perhaps in connection with implementation of one of the major, new
base-level systems discussed earlier.
The Committee's suggestions on measures to improve the management
of the base-level automation environment and the introduction of new appli-
cations software development and maintenance technology are provided in
Recommendation #4, Section IV.
-26-
Representative terms from entire chapter:
data processing
