Aging Military Electronics: What Can the Pentagon Do?

The long life span expected in current and future weapon systems, coupled with the fast-paced advances in commercial electronic technologies and the gradual erosion of the manufacturing base for military electronics have created the so-called "obsolescence" program that today plagues many weapon systems.

During the past decade, many acquisition reform initiatives at the Defense Department were motivated by the desire to tap the explosive growth in electronics capabilities of the commercial and consumer industry. In the 1980s, the Defense Department attempted to develop a new generation of unique, dedicated computers, only to find, after many years and more than a billion dollars devoted to the project, that the computers were obsolete and incompatible with the latest industry standards. This experience helped fuel the move to the use of commercial off-the-shelf technology in new systems and upgrades of aging systems.

It has become clear that the Defense Department is facing a significant obsolescence problem with respect to electronic components, exacerbated both by industry trends and by a reluctance to acknowledge its depth and breadth.

The evolution of the obsolescence problem is perhaps best understood by first reviewing the changing role of military electronics in the semiconductor industry during the past 25 years. During the 1970s, military requirements drove nearly all cutting-edge electronics research and development, and the military purchased about 35 percent of the industry's output of semiconductor components.

By 1984, the military was purchasing only 7 percent of the total domestic semiconductor output. But in spite of the reduced market share, military business was still desirable. The military still bought the most advanced and profitable chips and components, so most vendors continued to supply the military.

It was at this time, however, that the momentum began building to redesign military acquisition processes, in part to capitalize more effectively on the rapid developments in commercial electronics. This well-intentioned movement ultimately failed to anticipate the fallout from the electronics industry's explosive growth, which would significantly change the ground rules for manufacturers.

By the late 1990s, military purchasers confronted a commercial electronics base that was expanding exponentially. Yet there was a lag in the understanding of how commercial growth would affect the manufacture of electronics at the component level.

The military's share of component purchases is now under 1 percent. As a result, the military electronics market has become increasingly unattractive. Now semiconductor facilities cost billions of dollars to construct, and the only way to recoup these investments is by serving the mass market. To serve the military's needs involves low production volumes coupled with stringent manufacturing requirements. That is a sure recipe for un-profitability.

As the infrastructure available to support military electronics needs has eroded, the military is counting on legacy systems designed in the 1970s and 1980s to serve well into the new century. The U.S. Army's current roster of tanks and fighting vehicles is expected to be active until 2030, while the U.S. Air Force expects to use its current bomber fleet until 2040.

The severe obsolescence problems experienced by these aging systems can't be fixed by simple component replacement. Many components no longer are available. Those that remain available aren't competitive in either cost or performance with products routinely used in the commercial arena.

The use of components built to military specifications (Mil Spec) originally was driven by the need to deliver reliable weapon systems. Their life cycle, typically 10-20 years, corresponded with the anticipated life cycle of the systems in which they were installed. Their enhanced durability and long life-cycle offset their higher initial cost, making it prudent for designers to specify Mil Spec components.

It was expected that, by leveraging the rapid improvements and high volume seen in the commercial electronics market, lower costs would follow. However, in part because of the erosion of the military electronic manufacturing infrastructure, component costs often have grown.

Lower-than-expected savings from moving to commercial components is only one problem. A new set of increased costs, only now being fully understood, have actually driven the total life-cycle cost of commercial components higher than the Mil Spec components they replace. These additional costs include:

* Increased Testing Costs. The cost of ensuring that Mil Spec components met their specifications was borne by the manufacturer. A single thorough environmental qualification was viewed as a sound investment, enabling a given part to be used in multiple platforms and subsystems, across all services. Commercial component manufacturers understandably don't consider themselves a partner in this process. Therefore, each individual customer must finance the development of test vectors and equipment for the intended application.

* Reduced Life Cycle Costs. The commercial sector's pace of development makes their testing for extended life pointless. The peak availability and lowest pricing of a given component may last only six to 18 months, before the next generation of components takes its place. This short life can make it necessary to requalify new components several times during the service life of a given subsystem. It also means that multiple configurations of the system are deployed at the same time, driving up training and logistics costs.

* Proprietary System Architecture Cost. Many legacy systems were built using closed architectures, with proprietary form factors and interfaces. Therefore, integrating new components often involves non-recurring engineering costs. These costs are exacerbated, on a per-unit basis, by low production volumes.

* Assembly Priority and Schedule Risks. The subsystem and circuit card design work that is too costly for large defense companies to pursue in-house now often falls to small defense contractors. To cope with the unfavorable cost structure and low run volumes, some of these companies have formed alliances with commercial card assemblers. These commercial assemblers fund their cyclical process improvements based on serving the ever-evolving commercial requirements. The low-volume runs required for military electronics makes this prohibitive for most defense contractors.

These realities dictate that obsolescence be regarded as inevitable, and that its consequences must be mitigated.

One example of proactive obsolescence management is the work by the U.S. Air Force B-2 bomber program office.

The B-2 experience has demonstrated that there are no easy or comprehensive fixes to the obsolescence problem. The effort is labor intensive, must be based on a sound plan and can't be managed by simply dumping work and responsibility onto the prime contractor.

The B-2 process uses predictive tools such as the transitional analysis of component technology, known commercially as TACTech, to help define the obsolescence of each subsystem or line replaceable unit (LRU). The resulting analysis is further refined, by using the Pentagon's diminishing manufacturing sources management system (DMSMS) database to add information about component availability. This database is supported by the Defense Department's microelectronics activity, program offices and prime contractors. It includes information about last-time buys and the availability of specific components that may not be addressed by TACTech.

The B-2 team then develops options based on the TACTech/DMSMS analyses, assigns costs to each and makes recommendations.

Obsolescence management is primarily a tool for reducing or avoiding downstream costs, rather than generating immediate savings. However, the challenge can be addressed with a proactive, team-oriented approach, based on analyses using tools already available.

Donna Dillahunty, who works at the B-2 logistics management office, noted that, initially, "we never knew a part was obsolete until spares orders were returned from the vendor as no-bids. The real solution is to be proactive."

To determine where the potential problems were, she explained, "we first split the B-2 into its basic subsystems, and then divided each subsystem into its respective LRUs [line replaceable units]. We gathered engineering and cataloging information for every single component on every LRU, and entered it into our database, which is networked into a repository of obsolete parts information from almost every IC vendor in the world, along with lists of alternate parts that have been used by other weapon systems.

"Even though a part is obsolete, we may have enough on the shelf for the life of the weapon system, so it's not a problem for us. But more often than not, we don't have that luxury," she added.

The experience of the B-2 program office provides proof that it is possible to manage obsolescence challenges proactively. Nonetheless, to avoid dealing with these problems in a scattershot, isolated fashion, it is essential to build a military acquisition culture that routinely considers obsolescence issues as part of the decision-making process for each program.

Maj. Gen. John Caldwell, commander of the U.S. Army Tank-automotive Armament Command (TACOM), said that obsolescence is a "growing problem all of the Defense Department is facing. You can't avoid it or completely eliminate it. But you can work to mitigate the pain that it causes our managers who have to support old systems."

Success will come, he said, "when we make a significant team effort from both the government and industry to put in place the resources and processes to deal with this complex issue."

Recommendations

* Within each program, a process should be defined for dealing with obsolescence. The scope may range from continuous team reviews, for simpler programs, all the way to the complete management system exemplified by the B-2 program.

* Obsolescence mitigation should be established as a job-performance metric, especially for critical personnel such as engineers, logisticians, buyers and program managers. This serves notice that it is regarded as a key part of their jobs.

* Design engineers should use open systems and modular system architectures. This builds on the assumption that obsolescence is unavoidable and that at some future point, parts substitution and/or redesign will be necessary.

* The services need to ensure that there is adequate funding for obsolescence mitigation, both for legacy and new systems. Funds are required to support team activities and to underwrite recommendations, which may include parts substitutions, re-qualifications, redesigns or last-time buys.

* Government and industry managers need to monitor component usage patterns and maintain close vendor relationships. A recent example was seen when project managers, primes and subcontractors learned that the commercial cell phone market had locked up completely all supplies of a critical component thought to be readily available.

* Because the technical data for aging systems can be lost or become obsolete, it's essential to characterize and catalog component functionality for those components approaching obsolescence. With this functionality properly documented, it's possible to use custom or re-configurable devices as a design option to lower costs.

* A program to effectively screen and/or derate components needs to be established, with the cooperation of the original manufacturers. This can expand the available options when managers need to extend the life of LRUs and their incorporated components.

Philip Hamilton is vice president of Lau Defense Systems. Gorky Chin is vice president of Vista Controls, a Lau division that manufactures military electronics.

                Total Worldwide Semiconductor Usage Market
                 1984         1999
              (26.1B)     (149.5B)
Computer           39           50
Commercial         24           14
Communication      13           21
Industrial         11            8
Automotive          6            6
Military            7 [less than]1