Review of the Future of the U.S. Aerospace Infrastructure and Aerospace Engineering Disciplines to Meet the Needs of the Air Force and the Department of Defense (2001)

Government policies can mitigate or exacerbate the negative effects of budget cuts and program instabilities on the aerospace infrastructure. U.S. government and Department of Defense (DoD) policies have a double impact on the allocation and administration of program funds. First, they directly impact how much technical progress will be made for the taxpayer dollars spent. Second, they have an indirect impact on the quality of the defense aerospace technical work force—on both experience levels and the attractiveness of working in defense aerospace. Thus, policy makers must take into consideration the direct and indirect effects of policy and budget decisions.

The committee identified the effects of long acquisition times (i.e., the time between program initiation and initial fielding) on programs and evaluated the implications for the defense aerospace infrastructure. A 1986 Packard Commission report stated, “An unreasonably long acquisition cycle—10 to 15 years for our major weapon systems is a central problem from which most other acquisition problems stem” (CDM, 1986). The report noted that the long acquisition cycle also leads to unnecessarily high development costs (CDM, 1986). Today, more than 15 years later, the problem continues. In contrast, development cycle times in the commercial sector driven by competitive pressures have fallen dramatically. As of year 2000, the commercial aircraft industry had cut cycle time from 8 to 10 years to 5 years; the spacecraft industry had cut its development cycle time from 8 years to as short as 18 months (this reduction reflects, in many cases, modifications of existing platform designs as opposed to the development of new platforms); and the consumer electronics industry, from 2 years to 6 months. These reductions demonstrate that with a concentrated effort, cycle times can be cut in half or better (Hastings, 2000).

Long cycle times not only impact the quality of the products delivered to our warfighters (e.g., products are based on requirements more than 10 years old, and technologies are frequently obsolete and sometimes out of production), but also have a significant impact on the defense infrastructure that designs, develops, and produces these products.

First, the relationship between the development cost for a defense system and its development schedule is not linear (McNutt, 1998). Second, long acquisition cycles are major contributors to program instability, the most disruptive of which is the higher rate of cost increases. In 1996, the Air Force established the Lean Aerospace Initiative (LAI) to address this and related problems. Surveys of more than 100 government program managers and 80 contractor managers revealed that program cost growth attributable to budget changes alone averages just over 2 percent, year after year (LAI, 1996a, 1996b). The cumulative effect for an 11-year acquisition cycle is significant. In addition, cost increases in one program generally mean that budgets elsewhere must be reduced to make up the difference.

Cancellation is the ultimate form of instability, and the probability of cancellation for a defense program is just over 4 percent per year for each year the program is in development (Augustine, 1996). In a rapidly changing environment, programs based on 10- or 12-year-old requirements become increasingly vulnerable to cancellation. Unfortunately, programs canceled late in the development process result in zero return on very large investments. For example, four generations of Army air defense systems were terminated at an investment of $6.7 billion (Johnson, 1995). In addition to these huge losses, program cancellations are a major contributor to the difficulty of attracting qualified personnel.

Changes in program direction also occur frequently as the result of changes in leadership and changes in defense program annual guidance issued during the annual budget cycle. During an 11-year acquisition cycle, the leadership at each level, from the program director to the secretary of defense, typically turns over four to eight times; the most frequent

turnovers occur at the service and defense acquisition executive levels (Eash, 1998). These changes are reflected in shifting priorities and associated budget cuts and redirection (LAI, 1996a, 1996b). Budget cuts and program redirection often result in programs being stretched out as well. Of the top 20 acquisition programs now in place or planned, 6 had delays due to funding shortfalls, 5 had production cuts or stretches, and 3 had uncertain funding or were unfunded. The cost and commonality goals of the Joint Strike Fighter (JSF) have been characterized as unattainable. The F-22 production goal has been cut four times in eight years, unit cost is rising (partly as a result of production cuts), and congressional resistance to continuing the program is increasing. The production goal for the F/A-18E/F was cut by 45 percent in the Quadrennial Defense Review (QDR), and some Marine Corps resistance to continuing the program has been encountered. The Future Strike Aircraft (FSA) is currently unfunded and will not be produced prior to 2030. Numerous restructurings of the Comanche program have been forced by funding shortfalls. Production of the C-130J was delayed due to a funding shortfall. The KC-XX is as yet unfunded (Thompson, 2000).

The benefits of shorter acquisition cycles would be enormous. First, shorter development programs would allow individual workers to participate in more programs and thus increase the breadth and depth of their technical knowledge. Thus, working on Air Force programs would be more attractive. Second, shortening the acquisition cycle would save money that could be used to support more programs. Third, besides enhancing military capability, shorter cycles would reduce gaps between programs and reduce the loss of specifically trained, skilled, and experienced workers. Fourth, shorter cycle times would improve program stability by reducing cost overruns, cancellations, and changes in program direction. Reducing these disruptive changes in programs would lead to sizable cost savings. Improved stability in the workplace would make a career in aerospace more attractive. Finally, shorter cycle times would allow engineers to spend more of their time engineering and less on the busy work related to the acquisition process.

DoD acquisition policies have made it difficult for defense systems to take advantage of the latest commercial cost benefits and dynamic advances in commercial technologies. This has implications for the defense infrastructure. The size of the defense infrastructure depends on a combination of two factors—weapon subsystems and components that cannot employ commercial solutions and subsystems and components that could use them but do not.

As the use of commercial off-the-shelf (COTS) technologies and products increases, the need for defense-unique development capabilities decreases. A report by the RAND Corporation, Cheaper, Faster, Better? Commercial Approaches to Weapons Acquisition (Lorell et al., 2000), includes a case study showing the cost savings of inserting commercial-grade parts in the manufacture of military avionics components on higher volume, automated, dual-use production lines. The report cites the results of a U.S. Air Force program at Wright Laboratory for the development and manufacture of lower-cost modules for fighter and helicopter systems. Taking maximum advantage of commercial parts, the two modules were estimated to cost about 60 percent of the original F-22/RAH-66 baseline cost projection, even though the modules had not been designed for COTS insertion. The program did not permit basic electrical redesign of the modules. Partly because of this restriction, 10 percent of the parts remained military specification (Mil-Spec) and accounted for 50 percent of the module cost. The report concluded that even though commercial parts would have to be screened and possibly made more rugged or repackaged prior to use in military systems, they would often be less expensive than Mil-Spec parts (Lorell et al., 2000). In addition to the technical gains in performance, quality, and supply, the Air Force did not have to fund the development of commercial components.

The uniqueness of the defense infrastructure is related not only to unique technological requirements (e.g., operating in extreme conditions, such as high or low pressure, high or low temperature, the presence of nuclear radiation), but also to different design, development, and production processes and to all of the supporting disciplines. As the defense sector embraces a larger and larger share of commercial processes, the differences between the commercial and defense sectors will also diminish, and therefore the differences in the specialized training, skills, and experience of the work forces will diminish. Thus, the barriers that restrict movement of the work force between these sectors, during both peacetime and periods of crisis, will come down. In areas where the differences are small, commercial industry will provide an abundant source of skilled workers, and the issue of infrastructure maintenance will begin to recede. The challenge, therefore, is first to maximize the use of commercial solutions to satisfy military requirements and second, if commercial products are not acceptable, to adopt commercial processes and practices wherever practical in the development of defense-unique solutions.

Because many critical technologies and processes have few if any commercial analogues, the use of COTS products and processes can reduce the magnitude of the aerospace infrastructure issue but will not eliminate the need for a dedicated defense aerospace industry.

DoD recently commissioned the Defense Science Board (DSB) to review recommendations for financial policy changes in response to the declining financial health of the defense aerospace industry (DSB, 2000b). Based in part on

these recommendations, the Pentagon is implementing 27 regulatory and policy initiatives to improve the financial health and reflect the restructuring of the defense industry. The Pentagon will allow defense companies to keep part of the savings from efficiencies and reductions in overhead rather than continuing the current practice of returning the savings to the Pentagon. This will provide defense companies a greater opportunity to profit from R&D programs. Defense companies have traditionally taken greater financial risks in the R&D phase on the chance of larger profits in the production phase. The Pentagon also plans to increase multiyear contracts to stabilize production and lower prices. In addition, the Pentagon will increase its R&D spending by 10 percent annually over planned levels of $38 billion in FY02 and $37 billion in FY03 and FY04 (Capaccio, 2000).

In the last two years, DoD has begun to address the problem of acquisition cycle time in earnest. It has established a goal of reducing the average acquisition cycle time by 50 percent for all major defense acquisition programs (relative to a baseline of 11 years) started in FY99 or later (OUSD (AT&L), 2000). Reductions will be based not on reductions in the schedules for existing processes and activities, but on the development of a new way of doing business, including new processes, new activities, and new organizational structures and relationships. The objective is to develop a more effective, more efficient, and more time-sensitive approach. The focus on cycle time provides the vision and objective for implementation of a “lean” philosophy and serves as a metric of success in achieving that objective. Industry’s focus must be on improving design, development, and production processes and organizations. DoD’s focus must be on the establishment of policies and procedures that will enable this evolution.

The committee strongly supports this initiative. However, the committee cautions that extensive changes to procedures, concepts, and practices require careful deliberation. Changes should be made only after scrutiny by those who have had extensive experience in the processes and operations being revised. A headlong rush to embrace “faster, better, cheaper” can lead to some very undesirable and unexpected consequences.

The report on the National Aeronautics and Space Administration’s (NASA’s) failed Mars Climate Orbiter identified a “lack of identification of acceptable risk by the operations team in the context of the ‘Faster, Better, Cheaper’ philosophy” as a causative factor in problems with the program management (NASA, 2000). In short, changes made to improve cycle acquisition time require the attention and active participation of the highest and most experienced levels of senior management in DoD, the Air Force, and industry (Smith and Reinertsen, 1995).

Senior leadership in the Air Force has been visible in support of the cycle time reduction initiative. Both the Vice Chief of Staff and the Assistant Secretary of the Air Force for Acquisition have taken strong advocacy positions (Delaney, 1999; Lyles, 1999).

DoD has already taken some steps to shorten cycle times. For example, Instruction CLCSI 3170.01A of the chairman of the Joint Chiefs of Staff has been modified to suggest that time-phased requirements be used when feasible and to stress evolutionary acquisition and increased technical maturity before acquisition is initiated. The October 23, 2000, revision to DOD 5000–1 significantly emphasizes reducing cycle time in the following ways:

• A rapid, effective transition from science and technology (S&T) to product development;

• Using time-phased operational requirements;

• Requiring demonstration of a technology prior to the start of formal acquisition;

• Placing priority on evolutionary acquisition strategies based on time-phased requirements, proven technologies, and demonstrated manufacturing capabilities; and

• Initiating formal acquisition at or between any of the formal development milestones.

Continued efforts to reduce cycle time will be critical. Even after the current policy changes are fully implemented, much will remain to be done. At Toyota, the leader in implementing the lean philosophy, it took more than three decades for the reforms to permeate internal, supplier, and distributor systems. Toyota continues, even now, to find ways to make significant improvements (Womack and Jones, 1996).

Current DoD reforms are but first steps toward dramatically reducing development time lines, and they will contribute measurably to the efficiency of the defense industry. They will also allow greater flexibility in program design and lead to the development of processes more in tune with industry cycle times. Reform should be considered as a long-term undertaking, however, that will require continued visible support by Air Force leadership and civilian leaders in the Office of the Secretary of Defense.

Another aspect of the acquisition cycle that affects industry’s long-term viability and the health of the infrastructure is the distribution of work across five product (program) phases: (1) concept development, (2) demonstration and validation (essentially prototyping), (3) engineering and manufacturing development (EMD), (4) production, and (5) sustainment. Low industry profits, sometimes even losses, in the first two phases have historically been made up by higher profits in the EMD phase. Today, however, the amount of industry EMD has been reduced, which decreases overall profits.

Each phase of the product cycle requires different engineering skills. Indeed, if one phase were omitted for a period of time, such as the EMD phase on a new aircraft, the engi-

neering teams for that project would have to be disbanded. This would lead to increased costs, either to retain people during the hiatus or to reconstitute a team after the hiatus. Experience has shown that disbanding and reconstituting teams after a substantial period of time (e.g., five years) increases the cost of the subsequent EMD phase by about 35 percent. For a typical military EMD phase of $15 billion dollars, the increase would be $5 billion (M.Kennedy, RAND Corporation, personal communication, February 25, 2001). Industry is concerned that such reconstitution would be very long and expensive. The lesson for the Air Force is clear. The industry cannot survive by working in only one phase of the product cycle.

The aircraft engine component sector of industry has been a notable success. The committee reviewed this case in depth to determine if lessons for other industry sectors could be identified. The review revealed that a key factor in the success is the commonality between critical technologies in military and commercial engines combined with the military’s long-standing practice of investing in engine technology programs, such as the Integrated High Performance Turbine Engine Technology (IHPTET) Program. Stable investment has provided the continuity to sustain this sector.

Another aspect of the engine component sector that contributes to its success is the life-limited components and associated repair and replacement business. The lifetimes of engine parts are considerably shorter than the lifetime of the airframe. This creates a continuing revenue and profit stream for engine manufacturers on which investment strategies can be based for technology generation and development for product improvement and replacement. Involvement in the after-market business also enables the engine supplier to become actively involved with products that sustain engineering, thus evening out the workload between product launches.

In addition, the engine industry has benefited from a long-term technology investment program by the military. Investments by industry, the military, and NASA have fueled the technology initiatives in the IHPTET umbrella program on advances in engine technologies and meeting specific military requirements. In appropriate situations, the industry has applied advances in core engine technologies for commercial use. This dual-use arrangement has benefited the military by having the commercial experience feed back into the common military components and has benefited industry by sustaining it between military programs.

In summary, because a unique characteristic of the aircraft engine industry is that defense products can piggyback onto its commercial products, this sector is not a good model for the aerospace industry as a whole. However, this sector does illustrate the importance of sustainment. Long-term investment in technology programs such as IHPTET has enabled this sector to avoid the problems facing other industry sectors such as airframes.

The phases of the acquisition cycle taken together provide a business base that generates the profit and investment that enable firms to grow and prosper. For example, initiating prototype programs for the purpose of maintaining aerospace design teams but without robust EMD and production programs will not sustain the industry, although it might sustain the design teams that may be the “long pole” in reconstructing industry’s capability.

RAND has suggested a continuous program of three simultaneous military aircraft prototype programs, each lasting five years, which would require $1 billion per year in current dollars of dedicated funding and could mitigate the problem of loss of experience. Under the prototype plan, workers would gain experience on a new program every five years (Lorell, 1995; Lorell and Levaux, 1998). Similar programs could be used for aircraft engines and avionics. However, engineering talent is a rare commodity in today’s highly competitive technology marketplace, and firms are very reluctant to spend their scarce engineering resources for prototype programs without the prospect of profitability during production.

Although the committee recognizes and strongly supports the contributions that prototyping can make to attract, train and retain skilled professional designers, after considering alternatives, the committee concluded that continuing activities from R&D to EMD through production and product support are essential for maintaining all of the skills and team experience necessary for the aerospace industry to produce new aircraft. While prototyping was not addressed in depth by the committee, a separate study of the subject should be considered.

Another inhibiting factor in the expansion of aerospace markets is the restrictions imposed by licensing regulations, the International Traffic in Arms Regulations (ITAR), and the Arms Export Control Act. The U.S. government regulates arms exports to ensure that they do not adversely affect national security. The Arms Export Control Act authorizes the president to control the export of military items, which are licensed through the ITAR.

Nevertheless, exports are a major source of production, employment, revenue, and profits for the U.S. aerospace industry. The international component of sales for Lockheed Martin Aeronautics Company is now 70 percent; in the 1980s, international sales accounted for only 30 percent. In fact, international sales sustain the technical base and profits for investments in R&D and modernization (LMAC, 2000).

Foreign aerospace sales and cooperative projects produce a significant revenue stream that can be used to fund R&D.

For example, the Air Force is buying 30 F-16s from Lockheed Martin in the next five years, but Lockheed Martin has also negotiated significant foreign sales, including a large sale to the United Arab Emirates for the Block 60, a more capable F-16 model (LMAC, 2000). Current plans for an Air Force purchase of F-22s from Lockheed Martin call for 339 craft, with potential sales to some allied nations such as Australia (Wall, 2000). If Lockheed Martin wins the JSF contract, it plans to sell 2,800 JSF aircraft domestically, along with significant foreign sales, possibly a Harrier replacement for the British Royal Navy. Lockheed Martin has two significant cooperative programs with the Japanese for the F-2, a version of the F-16, and with the South Korean government for the KTX-2. Korea plans to build as many as 100 aircraft for its own use and to export the aircraft widely.

Export controls penalize American aerospace companies trading in the global market while providing little benefit to national security. In fact, they do the opposite by weakening the U.S. defense industry. Unnecessary regulations control many items that are readily available on the international market, such as advanced computers and encryption codes. In addition, the process by which these regulations are implemented is very complicated, time consuming, and uncertain; each expensive and burdensome paperwork and bureaucratic negotiation requires multiple permissions. The United States has two systems for controlling exports. The first, established under the jurisdiction of the U.S. Department of State, is designed to control the export of military products and technology. This system restricts exports unless there is a national foreign policy or security rationale for exporting the product. The second, established under the jurisdiction of the U.S. Department of Commerce, is designed to control the export of commercial and dual-use products, that is, products that are used in the commercial world but have military applications. Dual-use products are controlled if there is a security reason to do so; commercial products are controlled only where foreign policy sanctions have been invoked.

There is general agreement that this dual system is confusing at best and inefficient and counterproductive at worst. Many items are included in the State Department military product system simply because they have been modified for use on a military product, even if comparable commercial products and technology are widely available in the global marketplace. If such products must be controlled, they should be controlled as dual-use products by the Department of Commerce.

A major problem with the State Department licensing process is that only about 18 staff members in the Office of Defense Trade Controls (ODTC) have been responsible for processing approximately 45,000 licenses in the past several years. If a license is referred to DoD or other agencies for comment, the average licensing time is about 100 days. The problem was exacerbated in 1998 when Congress transferred commercial communications satellites and components from the Department of Commerce system to the State Department because of the alleged leakage of some rocket technology to China. This legislation has imposed additional restrictions, hampering the sale of communications satellite systems and services, sometimes putting U.S. industry, and therefore the Air Force itself, at a disadvantage (Douglass, 2000b).

For the launch of a U.S.-built satellite on a U.S.-built rocket from a U.S. launch facility for a foreign customer, the U.S. company must have an export license, which requires DoD and intelligence reviews and a technical assistance agreement from the State Department. In addition, a technology transfer control plan, an extensive plan that requires Department of State and DoD approval, must often be filed. If the product is valued at more than $50 million, congressional notification and approval are also required. Depending on the type of license, the entire process can take 10 months or longer. According to the Department of Commerce, exports of U.S. communications satellites and components fell by 40 percent as a result of satellites being placed on the State Department’s list of controlled munitions in 1999 (Reinsh, 2000).

After meetings between congressional representatives, including the Armed Services committee, the Foreign Relations committee, and the appropriation subcommittees, and industry associations, funding for ODTC has been increased; the review process for satellite exports to the North Atlantic Treaty Organization (NATO) and major non-NATO allies has been expedited; and export process reforms in the Defense Trade and Security Initiative have been ratified. These changes are expected to reduce the backlog, but more will be needed.

Despite export license reform, however, recent legislation, responsive to the alleged unlawful transfer of sensitive missile technology to China, has imposed additional restrictions hampering the sale of communication’s satellite systems and services, sometimes putting the U.S. satellite industry at a competitive disadvantage against European systems (Douglass, 2000b). The policy seems to be based on the assumption that the benefits of export sales accrue only to the buying countries and defense industry stockholders. This is simply not the case. Export sales of military items are an important element of the defense of the United States. The perils of exporting the wrong things must be weighed against the real benefit of exporting the right things. The State Department is not necessarily equipped to make this trade-off.

Although it may be difficult to strike a balance between fair trade and national security during periods of military and political change around the world, there are clear indications that the current balance is weighted in favor of restrictions. John W.Douglass, president of the Aerospace Industries Association of America, testified before the Senate Armed Services committee in March 2000, “German irritation with our licensing process is such that managers of DaimlerChrysler Aerospace, or DASA, Germany’s largest aerospace company, have been instructed to avoid purchas-

ing American components for defense and space products.” Douglass noted that DASA is in the process of merging with French and Spanish companies to form the fourth largest international aerospace and defense company. Restricting U.S. trade leaves room for new European and Asian entrants into component markets currently dominated by the United States (Douglass, 2000b).

Vigorous exports of U.S. aerospace products contributes to the financial health of the aerospace industry, it’s breadth and depth of work, the industry’s competitiveness in the global economy; and its ability to support national security interests. These national security issues should be given weight along with the security issues that result in extensive trade restrictions.

The term “test facilities” covers a very broad range of test capabilities used by the aerospace industry and government laboratories. The committee focused specifically on the test facility infrastructure of the “aviation” segment of the aerospace industry.

Full-scale, complete-airframe, static and fatigue test installations and other specialized test facilities can be very large and very complex. Nevertheless, they can be constructed and calibrated prior to the time they are actually needed for testing a particular aircraft design. This cannot be done for wind tunnels used for testing airframes and engines. Although computational fluid dynamics has made impressive strides in the last several decades and can define a best overall configuration, details must still be optimized. Fine-tuning engine inlets, exhaust systems, and wing-fuselage fillet configurations and minimizing interference shock effects at transonic speeds usually require testing in wind tunnels. Free-flight tests are usually more expensive and have a more limited range of test parameters than wind tunnels.

It takes years to build a large, well-equipped wind tunnel test facility and bring it to operational status. The long time interval is not dictated solely by design and construction requirements. Considerable time is also required for running extensive tests to calibrate a new facility. Because wind tunnel testing is required early in the design process of airframes and propulsion systems, facilities must be ready as soon as a decision is made to design and produce flying hardware. Wind tunnels are expensive to build, operate, and maintain.

The use of the nation’s wind tunnel facilities has declined significantly. As a result, some facilities have been closed or placed in an inactive state. Boeing is closing one of its tunnels, and several facilities at Arnold Engineering Development Center (AEDC) are becoming inactive because of lack of use (Boeing, 2000b; Heil, 2000). The major national ground test facilities used in DoD programs are at AEDC and NASA centers (Ames, Langley, and Glenn). Flight tests are conducted at a number of Air Force bases (AFBs), principally Edwards AFB and neighboring NASA-Dryden Flight Research Center. The viability and health of these facilities have been of much concern in the last decade or so because many of the major wind tunnel facilities have been in service for more than half a century. The alarm was raised by the failure in 1986 of the pressure shell of the 12-foot tunnel at the NASA-Ames Research Center. The facility has since been repaired, but it took many years before sufficient funds were made available.

A 1988 National Research Council (NRC) study of the NASA facilities, Review of Aeronautical Wind Tunnel Facilities, concluded that the NASA test facilities required serious immediate attention in terms of (1) maintenance and upgrading, (2) productivity enhancement, and (3) accommodation of new requirements (NRC, 1988). Although many facilities have been shut down in the intervening years, the more detailed recommendations of the study have generally been followed. The facilities that have been built since 1988 are primarily research facilities rather than testing facilities.

A similar NRC study of AEDC facilities, Future Aerospace Ground Test Facility Requirements, recommended that budgets for facility upgrades, maintenance, and repair be increased and called attention to the upkeep of Tunnel 16T, AEDC’s workhorse transonic testing facility, because of its importance to DoD programs (NRC, 1992).

Concerns over the viability of the major national low-speed and transonic test facilities led to a joint NASA-DoD initiative for a national wind tunnel complex. The project was originally envisioned as a two-tunnel complex that would cost about $2 billion. When the original cost estimates were shown to be too low, the plans were changed to one tunnel that could test for both speed ranges. The design for the tunnel was completed, but the program was dropped, and the tunnel was never built.

As the use of wind tunnel facilities has declined, interest in expanding the nation’s aerodynamic test facilities has also declined. The focus today is on reducing operating and maintenance costs of existing facilities. Putting some facilities in a standby, or mothballed, status and decommissioning— even demolishing—others are being considered. However, a number of significant issues must be resolved before any of these steps are taken. For a better understanding of these issues, several members of the committee spent a day at AEDC. Although the following discussion is based on the visit to AEDC, it is also applicable to other government test complexes such as those operated by NASA.

An NRC study completed in 1983, entitled The Influence of Computational Fluid Dynamics (CFD) on Experimental Aerospace Facilities, warned of excessive dependence on computations as a substitute for physical testing. The study concluded (NRC, 1983) that

This prediction has turned out to be correct. Although the performance of aerospace vehicles at design conditions can be reasonably simulated, performance near the operational limits (e.g., stalls, buffeting, drag rise) has to be verified by testing in ground-based facilities or in flight. Although the emphasis in testing has shifted, the DoD emphasis on meeting performance requirements has become more stringent, and as a result there has been no significant reduction in testing hours.

The amount of testing at a particular facility is generally used as a basis for deciding whether to continue to maintain the facility in an operational status. However, past use is not a good predictor of future need. Very little testing has been done in some AEDC facilities in the past three years. However, the decision of whether to mothball or decommission these facilities should be based on the need for their capabilities in the future. The Air Force should include an estimate of the test facilities in defining the development of future systems (Heil, 2000).

Large test complexes with a variety of individual facilities often share common support assets. A notable example is the common drive system at AEDC shared by the 16-foot transonic and the 16-foot supersonic tunnels. Other support, such as technician staff, instrumentation facilities, and model shops, is often shared among wind tunnels. Therefore, a decision to close only one test facility at a complex may not lead to significant total savings. Commonality should always be considered in the decision to deactivate a facility (Heil, 2000). Since industry is a user of these test facilities, it should also be a participant in the decision process.

Most wind tunnel testing is done under considerable time pressure. It is not uncommon for the design process of specific details of an air vehicle to be suspended until wind tunnel test results are available. Modifications and additions to basic test facilities, as well as changes to operating modes, are not always recorded in any formal way; they may exist only in the memories of the operating personnel. In this sense, test facilities are similar to manufacturing facilities in which critical so-called black-book knowledge is vital to efficient operations. Therefore, decisions to deactivate a facility should take into consideration the effects of losing the personnel who know how to reactivate and operate the facility (Heil, 2000).

The decline in test program activity in the United States has diminished incentives to modernize and improve test facilities. Most test centers have long lists of proposals for increasing the efficiency, lowering the cost, and expanding the envelope for testing and research. Because of budget limitations, most of these proposals have not been funded and are in an “on-hold” status.

For example, AEDC has a proposal to modify its Aerodynamic and Propulsion Test Unit cell to permit the testing of tactical missile systems and on-demand launch and recovery systems for low-cost access to space. “The Propulsion Wind Tunnel Cycle Time Reduction” proposal projects an increase of 25 testing days per year at a cost of $50 million. The modification would reduce the average time for installation or removal of a model from more than eight hours to four hours (AEDC, 2000).

These unfunded proposals have important implications for decisions to close or mothball facilities. AEDC has more than 60 defined, unfunded tasks. If the number of facilities is reduced, the effectiveness of the remaining facilities should be increased to offset the loss of test capacity (AEDC, 2000).

Before active test facilities are shut down, comparable lists of all government facilities should be reviewed to ensure that appropriate improvement and modification tasks of the remaining facilities are undertaken so as to guarantee maximum overall effectiveness.

Both NASA and AEDC are trying to reduce the operations and maintenance costs of existing facilities. Currently, NASA is changing to a full-cost accounting system. The cost at the NASA Ames Research Center, for example, would increase from $1,550 per hour plus the cost of power to $7,200 per hour plus the cost of power (NASA, 1999). AEDC is a government-owned, contractor-operated (GOCO) facility with cost procedure limitations that differ from NASA’s (Heil, 2000). Therefore, the cost of testing can differ greatly, and the facility with lower user costs will probably be used more, even if it has lesser capabilities than the more expensive facility.

Because the use of a facility is uneven, with considerable periods of inactivity, reducing costs can have unanticipated effects. Direct cost accounting has certain advantages from the standpoint of financial management; the long-term effect on a facility that is not used extensively is to increase user costs. If an organization cannot afford the increased costs, it will simply stop using the facility. As a result, costs

will go up for the remaining users, some of whom will then stop using the facility, and so on. If the facility is scarcely used, a decision may be made to close it altogether, even though the facility may be critical for testing future advanced systems. In the long term, new systems could entail higher system costs and perform less well than if they had been improved through testing in the facility. In conclusion, although they are expensive to maintain and operate and require continual investment to keep them efficient and up to date, major technical facilities, such as wind tunnels, will be needed in the future. Planning for future technical resources must include maintaining these facilities. Fiscal pressures on the testing agencies are intensified by the continued decline in the use of aeropropulsion test and wind tunnel facilities. These pressures could be alleviated somewhat if DoD and NASA agreed on which facilities will be needed. These facilities could then be properly maintained and upgraded, and excess facilities could be mothballed or closed. The cost of using comparable facilities should be as close to uniform as possible so that the choice of facility is based on availability of the required test capability rather than on price competition.

Facilities used solely for research could continue to operate at the discretion of the responsible agency. In that case, the Air Force would have to define its long-term system goals and translate them into specific requirements for test facilities.

As the commercial segment of the economy continues to increase, the burdens associated with defense contracts are becoming more difficult to justify and support. Because of their unwillingness to accept DoD acquisition rules, key commercial suppliers, such as Intel, no longer supply military-unique hardware as a matter of policy. Thus, the pool of available companies and technical talent from which DoD can draw is shrinking.

Federal Acquisition Regulations (FAR)—such as the requirement to apply government cost-accounting standards (CAS), the Truth in Negotiations Act (TINA) (Public Law 10 USC section 2306a), and the False Claims Act (Public Law 31 USC sections 3729–3733)—are frequently cited as major obstacles to efficient, effective relationships between the government and the industrial sector. Although everyone would agree with the idealistic intent of TINA, many of the regulations simply generate paperwork and increase administrative costs. Some progress is being made. For example, FARs do not permit companies to recover the full cost of benefits and incentives, such as moving allowances and certain stock options that companies may find necessary to offer to recruit new people in today’s marketplace (LMSS, 2000).

The DSB recently recommended making recruitment and retention bonuses for people with critical skills, such as software and avionics specialists, recoverable (DSB, 2000b). In October 2000, Undersecretary of Defense for Acquisition and Technology Jacques Gansler announced that this policy change would be implemented. As another example, in 2000, Congress passed an initial reform of the government-unique CAS; the threshold on the value of contracts governed by the CAS was raised significantly (Douglass, 2000b).

In FAR Section 845, “Other Transactions Authority,” Congress authorized the Defense Advanced Research Projects Agency (DARPA) to enter into agreements with private industry without adherence to the FAR. The intent was to allow DARPA to experiment in its contracts for R&D. Under Section 845, government organizations or agencies can establish relationships with industry much more akin to commercial relationships. Congress has since authorized the extension of Section 845 authority to the military services. In a report issued in 1999, however, the DSB concluded that application by DOD was still quite limited (DSB, 1999). DoD could use the opportunity to explore commercial-type arrangements with industry to evaluate a variety of alternative arrangements. Lessons learned could be captured and used to identify, evaluate, and provide support for meaningful refinements to FAR.