Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 THE LOGISTICS SYSTEMS MODERNIZATION PROGRAM: AN OPPORTUNITY FOR EXCEL1ENCE In Chapter 3 we described the organizational environment within which the Defense Logistics Agency (DLA) must function and the systemic changes that should occur in order for it to deliver the greatest benefit to the Department of Defense (DOD). In this chapter, we identify the major building blocks for planning and implementing the Logistics Systems Modernization Program (LSMP). We believe that the LSMP can be the DLA's blueprint for excellence by providing the driving force for deploying and executing its modernization plans. The LSMP should be built on a solid planning foundation that is long-term in vision and evolutionary in its implementation. This planning should start with a set of goals, objectives, and critical success factors. These and other considerations needed to establish a foundation for the LSMP are treated below in the section "Building the Foundation." Next, we address the management and organization of the LSMP. A major function of the DLA is materiel supply and we discuss a number of elements of integrated materiel management. Contract management is a major part of the DLA's mission. About 35 percent of the DLA's work force is engaged in contract administration services. This major business area is treated below in the section "Contract Administration." Finally, a range of technology issues is discussed. BUILDING THE FOUNDATION Over the past year, the LSMP has evolved and is beginning to coalesce from a fragmented search for the future to a comprehensive, unified program to drive the modernization of the DLA. The building blocks for such a plan, while not fully defined, have been identified and placed within the context of a program schedule and resource management plan. Goals, Ob; ectives, and Critical Success Factors We believe that a broad vision of where the agency is going should be translated to a top-down, long-term view of what it wants to achieve through the LSMP. We have noted that this has been a missing ingredient 11
12 in the planning process for the LSMP, and the agency has responded to our observation. The goals and objectives recently set forth by the agency (private communication to the committee by General Vincent Russo, November 19, 1987, on file) are a good start toward providing the necessary top-down direction. We feel that strong, clear, and concise goals are needed to address the future operating environment for the DLA. Based on the goals, objectives need to be developed and refined to deal with more immediate concerns. Specific guidance is needed for the management and enhancement of the DLA's major functional areas: Materiel management and contract administration. Critical success factors have been recently identified by the agency. These will help avoid problems and further focus the direction of the LSMP (private communication to the Committee by Jean Lakey, January 25, 1988, on file). The goals, objectives, and critical success factors should be considered a "living" document that can change and expand as the agency moves into the future. It is evident from our most recent reviews that the agency recognizes this, and that those now working on the LSMP have a better vision of how to direct their efforts and gauge their progress. We believe it is important for this to continue. Business Area Analysis As part of the LSMP planning process, the DLA made a great investment in its future by putting forth the effort and resources (time and people) to define the agency's businesses from the bottom-up. This process included an analysis of the agency's information requirements, which it first developed along its vertical organizations. These results are being further subjected to cross-functional analysis. Once this effort is fully extended to the logistics agencies of the services, the DLA will have the basis for translating information requirements into technical architectures. This work should also provide an extensive basis with which to evaluate organizational structure and operational processes. It has the potential to identify a new operating environment with new structures that allow new and more efficient ways of doing business. At this juncture in the business area analysis (BAA) development process, the DLA should concern itself with what has been a problem with past BAA-like efforts. In too many cases, after completing the rigorous data collection effort, those directly involved become exhausted or burned out, and with leadership refocusing on new issues, the process ends without realizing its full benefit. The DLA is now in the most critical and productive phase of the BAA and should maintain the momentum of the process. 1. Management-level guidelines that are key to a successful modernization program. Uses lessons learned in previous programs.
13 The Logistics Systems Modernization Program Blueprint The LSMP blueprint, or overall plan, is not fully developed, but the building blocks are emerging. A new business model is being developed by the DLA that identifies the functions it performs. These business functions will then be related to organization, controls, markets, and resources. This model, along with a new understanding of how the DLA must operate in an integrated partnership with industry and its customers, should motivate efforts to: resolve regulatory constraints; effect new policies and procedures to foster the new integrated relationships; revise budget preparations; and facilitate the exploitation of technology to fulfill management expectations. We believe that the LSMP should focus on both internal and external changes that will allow the DLA to meet increasing demands from its customers both efficiently and reliably. A major building block in any large development process is the time schedule, including defined and measurable milestones. For the LSMP, this includes events, agency responsibility, and the time phasing of each element of the development schedule. The DLA has established a program management system that provides a program schedule. What is needed now is a critical review of each scheduled activity and its sequence to determine if overall schedule improvement is possible. Additional program management tools such as program evaluation review technique (PERT), critical path method (CPM), or earned value are advised as the program progresses. MANAGEMENT AND ORGANIZATION The scale and complexity of the LSMP requires strong management and organizational support. The formation of the DLA from various 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. We believe that the agency's vertically organized and unconnected ("stovepipe") operations could benefit from increased horizontal integration. The LSMP program offers an opportunity to achieve integration across activities within the DLA. In order to accomplish such internal integration, management must have the authority and responsibility to enforce the accomplishment of program objectives across the diverse and autonomous organizations and activities within the DLA. We believe this can best be achieved through a program management office. However, while the program management office should be able to control development of systems, it will take strong guidance, attention, and support from top management to enforce integration across activities and make the LSMP work. In addition, this opportunity to integrate across functions can only be done with an organizational viewpoint and degree of cooperation that comes from senior management direction.
14 Program Management Office We have consistently believed that the scale of effort dictated by the LSMP requires the full attention of a management group whose sole responsibility is the successful achievement of modernization. We believe that this is best accomplished through a program management office (PMO), organized along the lines employed by the services. Such an office would provide the needed continuity, serve to focus the agency's resources, consolidate its varied requirements, and plan for an evolutionary set of enhancements to meet short-term and long-term goals while maximizing the useful life of the investments made along the way. We strongly agree with the DLA's creation of a permanent program office for the LSMP and recommend that it continue to be strengthened through the addition of staff with managerial and technical skills, by training and developing present staff, and by assuring the tools and methods are sufficient for the task. In doing so, the DLA will need to move away from its traditional approach of decentralized management that places most development in the field activities. Strengthening the PMO and helping it exert greater and more centralized control over the tasks it delegates may be unnatural in a highly matrix-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. One way to reduce the problems caused by matrix operations could entail assigning functional and technical people to the program office for extended periods. During that time they would work only for the program office and be evaluated by the program office. In such a case, we believe that prudent and seasoned professionals will be responsive to the program while keeping their home organizations informed. Personnel Skills Obtaining individuals with sufficient expertise in areas such as the information and management sciences is a major problem for the government and not just the DLA. We believe that such skills are essential for management of the LSMP program, development of its systems, and operation and maintenance of the information systems needed in the DLA. We expect that contractor help can be used effectively to augment existing agency resources. However, the DLA should take immediate steps to obtain or develop the skill levels it needs to plan, direct, and manage the acquisition of automation systems. Automation Systems Development The scope of the LSMP project is extensive and will require the development of a large number of systems. We believe that the LSMP is too large and encompasses too many diverse functions to be managed without subdivision into major components. We caution against the concurrent
15 modernization of the DLA's functions. We prefer to see the DLA start with the most pressing and welI-defined business areas that are critical to its mission. We suggest that the DLA view the components of the 'LSMP as a number of subprojects. The subprojects should be planned as components of the two major DLA functions, materiel management and contract administration. The development of the LSMP 'should begin with designing a technical architecture framework. Starting from this framework, new systems can be developed sequentially and well ahead of the end date for the LSMP. A new technical facility will provide an early-payoff from the program by improving performance, information exchange, and reliability. Sequential additions and modernizations should also reduce the risks associated with integrating a system whose pieces must come together in a short period of time. A base architectural framework will also improve the likelihood that the early components will tee' compatible with the complete system. Decision support systems, discussed at greater length in the next section, are an important aspect of the LSMP and their development requires careful attention. These systems cut across all of the components of the LSMP and will serve many agency personnel from the director down to the warehouse worker. If properly designed, they will provide timely and accurate information in a form that is directly usable for the decisions being made. Decision Support Systems We did not find that the DLA had the capability easily to ' locale, access, format, and display 'information that decision makers need for many routine decisions, or that management needs to make timely, substantiated, and objective assessments of agency performance. The DLA needs to enhance significantly the ability of its information systems to support decisions. Decision support systems should draw information from certified agency data bases and provide it where and when needed. Such systems are commonly used throughout the commercial sector and do not pose inordinate technological problems. The LSMP provides a unique opportunity for the DLA to plan for and incorporate this capability in its operations. The systems need not, and should not, rely on a single data base, but the location of the inflation should be transparent to the user. While such capability should be available to the key decision makers throughout the agency, some security measures appear to be in order and need to be considered in developing the requirements for this function. - ' Decision support requirements should be a part of the LSMP. The business area analysis process should identify the information needed to support day-to-day decisions. In addition, we strongly urge the agency to 2. Data bases that are considered to be definitive by an organization and on which it intends to base its operations.
16 implement a prototype decision support system for use by the director and his staff. This will provide an excellent test-bed with which to evaluate and learn about such systems. It will also serve to involve senior managers in its design. INTEGRATED MATERIEL MANAGEMENT A fundamental service provided by the DLA is the timely and cost-effective supply of commodities and spare parts to the services.-This mission entails several activities including forecasting demand, cost effective procurement, and the maintenance of the lowest possible inventories consistent with the timely supply of the services's needs, within the framework of the Office of the Secretary of Defense (OSD) policies and defense guidance. The LSMP provides an opportunity to improve the DLA's performance in each of these activities. We believe that exploitation of these possibilities will improve readiness, reduce inventories, and increase credibility in the budget review process. An outline of an updated, efficient, and integrated materiel management plan should be in place before developing a long-range management information system (MIS) plan. Requirements Determination The DLA needs to forecast demands for support of the services. Improved ability to forecast requirements would enhance the DLA's ability to insure that investments in inventories are keyed to appropriate priority and readiness goals. However, random equipment failures and unanticipated and unprogrammed military actions both contribute-to the uncertainty in demand. In order to improve the accuracy of its forecasts, which would permit a reduction in inventory animal improvement in readiness, the DLA needs to know which end-items go with which weapon systems. Additional information that is needed regarding weapon system and end-item relationships includes item criticality, weapon system and end item inventory data, current and planned force deployment schedules, and operational activity levels. Procurement The DLA was established to take advantage of the economies of scale that are inherently present in procurement of similar items. The economic advantage of this leverage to secure attractive pricing, and to maintain effective vendor relationships, is real. However, the DLA's current procurement practices do not adequately or systematically exploit the advantage of scale economies in procurement nor do they call for sufficient monitoring and record keeping. For example, DLA decisions on order quantities are made by inventory managers without information on price reductions that might ensue from larger buys.
17 The Defense General Supply Center (DGSC), one of the DLA's warehousing and supply operations, invites potential contractors to furnish additional quantity and price data for all solicitations of $25,000 or more, which accounts for about 60 percent of the dollars spent. The DGSC does this by reference to Federal Acquisition Requirements (FARE in the solicitation. If this data is provided it is forwarded to the inventory manager for consideration on future buy quantities. For small purchases of less than $25,000, which account for 40 percent of the dollars spent, price-break information is requested. However, if furnished, this data is only used for that procurement. Thus, there appears to be significant opportunity for savings. We suggest that procedures bee established-for routinely collecting and reporting price-quantity schedules to inventory managers on a regular basis, perhaps in a common 'date base shared by buyers and inventory managers. ~ ' Moreover, inventory managers are not provided with tools to permit them to calculate economic procurement quantities that effectively trade off such scale economies in procurement and storage costs, even though methods for doing so are available in the operations research literature. Inventory managers should be provided with modern operations research - tools. In addition, consideration should be given to providing inventory managers with information on the time dependence of item pricing resulting from seasonality or opportunities for special pricing since such information might influence inventory managers to buy at favorable times. In that event, inventory managers should be provided with operations research tools to help them evaluate such opportunities. In reviewing the procurement process, it was noted that procurement administrative lead time (PALT) has increased dramatically in recent years, reportedly because of the increased use of competitive procurements. We believe that the development and execution of the LSMP provides the DLA several opportunities to reduce the PALT. For example, immediate accessibility to the technical data base, aggregate contractor profiles, common data base elements to serve both the inventory managers and the buyers, and increased accessibility to an expanded and concurrent contract administration data base all have great promise in reducing PALT without diminishing the emphasis on competitive procurements. Inventory Management Establishing the most effective and efficient methods for managing stocks under the responsibility of the DLA, including how they should be distributed among wholesale depots, depends on many variables. _ ~ ~ determination process, the weapon application, item criticality, the end-item population, and location of forces affect not only the quantity to be procured but also location and how effectively such items must be managed. We have not attempted to diagnose how effectively the DLA is coordinating all of these considerations. However, we suspect that there the various locations (e.g., at vendor plant, DLA or services' stock points),-is a complex matter that ~ ~~ As noted above in the reauirement~
18 may be significant opportunities for cost-effective benefits by reevaluating selective policy and practices. The most apparent examples are addressed below. We found that the DLA needs better visibility of the materiel assets it is responsible to supply, if optimum application of these assets is to be expected. Currently, the DLA does not have visibility of retail stocks held by the services. Although we are not sufficiently knowledgeable of the cost of expanded asset reporting, it is apparent that given a policy of expanded visibility at the wholesale level, the DLA could frequently defer or avoid additional buys. Such visibility could also;enhance readiness by providing faster response to priority requirements in a shortage environment. Considering that one-half of the DLA's reported work load at the receiving dock is for returned materiel, we anticipate a significant overall reduction in inventory investment if excess stocks are considered as wholesale assets as they become apparent at the retail level, when compared to the authorized retail allowance. However, there may be situations and circumstances where more visibility is not feasible and a revised credit policy could motivate better use of assets. We believe it would be beneficial if all excess property held by, or in transit to, the Defense Reutilization and Marketing Office (DRMO) were made visible and readily accessible to the inventory managers. This enhancement is believed to be technically reasonable and feasible in development of the LSMP. The determination of U.S. industry to produce internationally competitive products has spurred creative, new approaches to old problems. The emerging techniques of U.S. industry, where modern procurement, inventory, and manufacturing processes are being implemented, clearly show the need to link information subsystems. Many concepts that are being used to modernize the U.S. production base have the potential to significantly improve the way the DLA does business. One new concept is just-in-time (JIT) inventory which seeks to improve component quality while reducing inventory. However, this concept is not compatible with long standing, defense-wide practices of separately ordering, auditing, paying, storing, and distributing end-items. The defense logistics system practice is to hold sufficient stock in its inventory to meet its end-user demands. This practice of defense-held, extensive inventories needs to be reexamined in light of emerging techniques that guarantee delivery of fresh, updated materiel while minimizing the inventories needed to support operations. 3. Current information on the location, availability' and quantity.
19 Data Base Management Throughout the briefings and particularly from examining the various BAA reports, we became convinced that the system modernization inherent in the LSMP concept affords the DLA an unparalleled opportunity to create accessible data bases consisting of data elements needed to serve many different but related functions. We suggest that the DLA consider categorizing its data bases into three groups--technical, operational, and management. The technical data base includes: provisioning, cataloging, drawings and other technical type data included in bid sets, and the approved sources when applicable. The operational data base includes: inventory managers' demand history, assets on hand, under-contract quantities and on-order quantities not awarded, in-transit, applicable portions of contract files, and excess assets outside the normal reporting system. Data normally maintained for use by buyers in the procurement process are considered to be in the operational category. Customer orders in process at the inventory control points or storage depots are also included. The management data base includes those elements of data that account for budgets, fund expenditures, performance appraisals, and staffing. Surge Capacities for Contingencies and Mobilization In an attempt to find more economical stocking concepts, one must weigh the degree to which it seems wise to allow vendors to carry inventories for the services. The quality of response desired, including the tolerable percentage of stock outages and delays to be allowed and the ability to be responsive to surge demands during a contingency or mobilization situation, require careful consideration. During the initial visits and briefings, we found that opportunities presented by the LSMP for improving peacetime efficiencies were being fully considered, but we did not find that sufficient consideration was being given to factors associated with designing, structuring, or sizing for more capacities to meet increased volumes during contingencies or mobilization. We believe that such needs should be carefully analyzed and closely associated with the LSMP effort and supporting justification thereof. CONTRACT ADMINISTRATION Three primary functions of a supply operation are: determination of requirements, contracting for those requirements, and administration of contracts. Irrespective of their relative importance or the scope or magnitude of the duties that would be included in each of the three functions, there is a clear and essential linkage between them. The objective of logistically supporting end-users should include excellence in each area. While materiel management is often identified as the determination of
20 requirements and distribution of assets, modernization of that process alone will be of little value unless the total logistics system embraces the contracting (procurement).and contract administration functions. This section addresses the principal contract administration functional areas that must be.upgraded and integrated into the agency's overall plans for achieving a modern, effective, and efficient logistics system. - Contractor Profiles The basic role of the contract administration office is to ensure timely delivery of quality products to defense users of those products. In performing this role, it is essential that the contract administration office have sufficient, current, accurate, and complete.information on the performance capabilities of the suppliers with whom it deals. Such information, referred to as contractor profiles, is also needed in the buying offices for determining potential sources of supply for additional products. The DLA's contractor surveillance system does not provide complete contractor profiles for the firms under its administrative overview, even though such information is needed by the procurement and contracting officers. The contractor profiles that are produced are only available on a limited, regional, plant-by-plant basis. Furthermore, this fragmented information is not merged into an overall, current, and accurate view of a contractor's defense contracts, or the effectiveness of the contractor's performance on them. Complete and current profiles could be produced by a modern, electronically -linked system. Such a system should cost less and provide far better results than the labor-intensive, repetitive assessments of fragmented data that the DLA currently must perform. Linkages with Contractor Data Bases In FY-87 the DLA supply centers bought and sold about $12.4 billion of fuels, subsistence, clothing, medical supplies, and hardware. Imbedded in the cost of these supplies are the capital expenditures made by defense suppliers for their own information-generating systems. Yet, there is no forum for coordinating government and industry information systems. To the extent that the defense supplier base is contractually supporting and servicing the agency's needs and generating the data necessary to fulfill contractual requirements, and to the extent that the agency itself is Generating data for its needs, it would be in the interest of both to establish compatible electronic linkages that reduce the cost of both, and thus those of the DOD as a whole. Any long-range view of a modernized logistics management process for the DLA should encompass the desirability of such linkages.
21 Financial Management in the Acquisition Process Precise categorization of funds, reliable budgets, correct appropriation and accounting data, bulk and line-item contract funding, earned discounts, and payments to contractors are included in the stream of financial transactions that the DLA handles. Doing so efficiently and quickly is critical to the DLA fulfilling its fiduciary obligations to the government and also to the financial viability of its contractors. Institutional demands for acceptable accuracy in this functional area have not found the DLA wanting. However, in its long-range modernization program, the DLA should strive for more efficient and more economical handling of the continuum of budgeting, accounting, and financial data. In addition, that data should be accessible to, with security considerations, and compatible with other functional area data emerging in electronic form. The accuracy and timeliness of accounting information and payment for products and services rendered is essential to DLA and service-wide credibility and motivation of its supplier base. The LSMP should emphasize the need for a modern, accurate, responsive, and integrated financial management and accounting system for all phases of the acquisition logistics process. Contract Audit ing The contract auditing function is performed by both the DLA and the Defense Contract Audit Agency (DCAA). Based on briefings on the subject, the two agencies must work together in greater harmony and with less overlap. No evidence was found indicating that the two agencies have a plan for cooperating with each other or have been given OSD guidance on distinguishing their respective responsibilities. The result is an undesirable overlap of responsibilities and confusion within the agencies and the contractor community. This condition is neither supportive of the DOD's overall mission or the purposes of either agency. While the difficulties of obtaining a more cooperative relationship between the two agencies is recognized to be much broader than the compatibility of information systems, such compatibility would provide ~ common communication medium and facilitate the exchange of data needed by each agency and could lead to better coordination. TECHNOLOGY ISSUES Choices of hardware, software, and data bases are major considerations facing the agency as it develops and refines its automation system plans, but there are broader issues that have a more far-reaching effect on the agency. The following sections identify and discuss what we believe are the major technological issues that the DLA must deal with as it moves forward with its modernization plans.
22 Distributed Versus Centralized Architecture The current DLA system architecture is a hybrid configuration of multiple and autonomous computer facilities that cannot be interconnected and that cannot share data through electronic exchange. Further, since the applications do not work from a common or shared data base, on-line searches and exchange of data across centers are not possible. 4 While much of the literature on distributed data processing exaggerates the current capability of that technology, a great deal can, nonetheless, be accomplished in a well-designed system that has been built expressly for distributed processing. Additionally, the technology is improving rapidly, and within the time frames for the LSMP, we expect that a distributed processing system that is both robust and data-base driven will provide a cost effective capability when compared to the centralized architectural alternatives. Considering the DLA's dispersed operations and the applicability of distributed processing, we recommend that the LSMP plan on capitalizing on this technology when it develops its technical architecture requirements. Standards Establishing information system standards early in the LSMP development cycle is vital to the successful integration of information within the agency and to sharing information across the services. Such standards also affect portability and the cost of software maintenance. Standards apply to the near-term improvement projects known as critical baseline enhancements (CBEs) and their transition to the LSMP architecture. The standards that should be established for the LSMP include those that deal with network communications, hardware architecture, operating systems, programming languages, and data bases. The OSD has provided policy guidance (memorandum from Joseph Wright, June 22, 1987, on file; and memorandum from Donald Latham, July 2, 1987, on file) on the use of Open System Interconnection (OSI) standards (U.S. Government Open Systems Interconnection User's Committee, 1987), and the DLA should adopt these while posturing itself for new DOD initiatives such as the Computer-Aided Acquisition and Logistics Support (CALS). Since OSI standards are not currently implemented in all vendor equipment, the DLA should incorporate de facto standards such as those employed for the Defense Data Network (e.g., TCP/IP protocol). 4. Data and processing that occur in more than one place. Generally supporting the needs of an organizational element located there. 5. Accurate, current, usable, and complete. Source transparent to user.
23 An Open System Architecture (OSA) will permit the DLA to use the useful products of many suppliers and avoid being locked into proprietary, single-source designs. OSA is an effective means to preserve and protect the automation system investment. The LSMP should be positioned, through practical software and hardware decisions, to support this concept in the future. The extent to which open systems architecture is achieved will determine the degree of freedom that the agency will have in exploiting available commercial software and incorporating new hardware without paying high reprogramming costs. In the area off data bases, the DLA should choose a standardized data-base management systemic) and standardize on formats and structures for the data-base entities that are used by its different activities. We suggest that data-base design and implementation may be a critical path item that could seriously impact subsequent development efforts. Overall, we believe that the DLA should accelerate their investigation of the standardization alternatives and set schedule dates for selection of appropriate standards in each relevant technology area. Security Security issues are a major consideration for the LSMP. Security refers to the protection of information from unauthorized access, disclosure, alteration, and destruction. Protection may be required while the information resides in computer systems and while it is being transmitted over communications systems. For information bearing national security classification, well-developed procedures exist for its storage and transmittal in hard-copy form. Trusted computer systems are in the process of being characterized and certified (U.S. Department of Defense, Computer Security Center, 1983~. Trusted computer networks are more difficult to specify, but the effort is being made (U.S. National Computer Security Center, 1987~. In general, stand-alone computer systems are much easier to keep secure than are interconnected ones. End-to-end telecommunications security is hard to achieve and rests mainly on the development of third-generation terminals capable of encrypting digital data and digitally encoded voice. Some unclassified but sensitive information requires protection for proprietary and procedural reasons. In response to this awareness, President Reagan in 1984 issued National Security Decision Directive 145 (NSDD 145~. Among other things, the directive established responsibilities for providing telecommunications- and automated-information systems security guidance to the departments and agencies of government. Measures for achieving such protection are not well standardized. Public Law 100-235 (U.S. Congress 1988) recognizes the problem for government agencies and gives responsibility to the National Bureau of Standards (NBS) in coordination with the National Security Agency (NSA) for developing computer security standards and training for the civilian federal computer systems and managers. The intent of the new law was to arrange that a nonmilitary agency, NBS, would relate to nonmilitary systems, while leaving a military agency, NSA, to relate to
24 military systems. The new law does not change the relationship between the DOD, NSA, and the National Computer Security Center. Thus, the DLA would look primarily to the NSA for technical assistance. The available technology, as just noted, is limited in its effectiveness and can quickly accelerate a system's cost. Employing security measures without care for compatibility and transparency can restrict operations and severely limit hardware and software options. Sharing of data would be reduced and maintenance costs would be greatly increased. The use of security measures should be based on advice and assistance from the OSD and the NSA. In general, we suggest that security measures be evaluated based on the value of what is being protected, the severity of the threat against it, and, hence, the risks associated with its misuse. This will also require balancing of the cost of security at several levels of implementation against the probable consequences of compromise. We also expect that some operating efficiencies will be sacrificed because of restrictions in the distribution and common use of information. While it may be difficult to quantify this reduced level of integration, it should be included as one of the cost components of security. The DLA is investigating security requirements in the light of current policy. The agency should expedite this process and conduct a cost-benefit analysis to determine the trade-offs and practicality of imposing security requirements. Since security will have such a major impact, it should be dealt with as soon as possible and not be deferred. Information- and Knowledge-Based Systems The DLA is placing heavy emphasis on the development of knowledge-based systems in key areas. We strongly support the exploitation of this technology and believe that it can provide significant assistance in a number of areas such as procurement, transportation, and contract administration. However, we believe that the technology is often oversold and caution that it not be viewed as a panacea. Certain conditions must exist for knowledge-based systems to be applicable to a situation. We recommend that the DLA continue to use expert outside advice to carefully analyze each instance where artificial intelligence (AI) is to be employed to determine if these-conditions exist. A small application should be selected as a test-bed in order to provide the agency with an opportunity to gain experience for larger and more pervasive applications of this technology in an evolutionary manner. Because this technology is relatively new, contractors have often overstated its capability and their own. It is important to verify that contractors who promise to deliver the technology have the necessary expertise in the area. Logistics Research It has long been accepted that a DOD commitment to support and fund both basic and applied research in areas related to weapons systems development is essential to the maintenance of the defense capability of the free
25 world. In spite of the fact that weapons systems cannot be delivered without adequate logistics support, there has been almost no continuing commitment by the DOD in recent years to long-term basic and applied research in logistics. We believe that this deficiency is costly and that it needs to be corrected. Major advances are taking place in many areas of logistics research, largely in the commercial world, including, for example, multi-echelon inventory control, expert systems, distribution planning, materiel handling, routing, large-scale optimization, and optimization under uncertainty. Because of the DOD's minimal support for basic and applied research in these areas, the DOD is missing out on the great potential that these areas offer for improving readiness while at the same time substantially reducing massive DOD investments in inventory. Furthermore, the DOD places itself at risk when it tries to infuse new technologies rapidly into an unprepared environment. For these reasons, we strongly endorse the DLA's initiative to recognize and fund logistics research. REFERENCES U.S. Congress. Public Law 100-235, January 8, 1988. U.S. Department of Defense, Computer Security Center, 1983. Department of Defense Trusted Computer Systems, CSC-STD-001-83. Fort George G. Meade, Maryland. August 15, 1983. National Security Decision Directive 145 (Unclassified Version), September 17, 1984 U.S. Government Open Systems Interconnection User's Committee. 1987. U.S. Government Open Systems Interconnection Profile. U.S. National Computer Security Center, 1987. Trusted Network Interpretation of the Trusted Computer System Evaluation Criteria. NCSC-TG-005, Version 1. Fort George C. Meade, Maryland. July 31, 1987.
