Note: For the purposes of this article, “military system” refers to a complex piece of defense equipment, such as a warship, tank, airplane, communications system, etc.
Procurement and development costs have become a major issue for militaries around the world in the past few decades. One need look no further than the largest national newspapers to see examples of this: rising costs for programs such as the F-35 often make headlines and cause people to shake their heads in disbelief. Many are confused. Why is it so hard to deliver military systems on time and under budget? What makes military procurement so prone to overruns when many private-sector companies are able to churn out new phones and cars almost yearly?
One thing that is crucial to understand is that military programs are far and away more complex than anything a civilian would ever purchase. Aircraft such as the F-35 incorporate millions of lines of code, cutting edge composite materials, stealth coatings, and innumerable sub-components. Military programs are often on the bleeding edge of high technology, meaning that extensive research and development is needed and much of the requisite technology is completely untested.
The distribution of the research and development costs in military procurement is also more troublesome. In a civilian consumer goods, R&D costs are distributed over a vast number of units. In simplified terms: if a phone costs one billion dollars to develop and one hundred million sets are sold, each consumer is only shouldering $10 in R&D costs. If a submarine costs five billion dollars to develop and only fifty are built, for each submarine purchased $100,000,000 in R&D must also be accounted for, meaning the overall cost of military systems is often much higher than the cost of an individual unit. The necessity of paying R&D is also complicated when the number of overall units is cut, leaving the same amount of R&D to be spread over fewer units (which can cause unit costs to become unacceptably high).
When a new military system is designed, engineers have to grapple with fitting a massive amount of new technology onto an untested platform. The potential for issues to arise and delays to surface is referred to as “risk.” Private industry is able to easily test and incrementally improve upon products because the overall cost of designing consumer goods is relatively low compared to defense products. This allows private industry to quickly develop and integrate new technologies with relatively little risk. Essentially, private industry has the luxury of trial-and-error when designing products, because the failure of a new technology is acceptable when development costs are low.
Defense manufacturers do not have the luxury of frequently redesigning their products and testing new ideas, as the defense market is not large enough and one failed design can bankrupt a defense firm. Instead, there are often many decades between each clean slate design, and one system may serve for many decades before replacement. Thus, instead of gradual low-risk improvements to a product, defense engineers must often implement many untested elements in a new system. Adding so many improvements to a system is highly risky, as each new element has the potential to cause technical issues if it does not function correctly.
Military systems are hugely complex and labor-intensive. For example, designing the Virginia-class attack submarine took 35,000,000 man-hours. Each individual Virginia-class takes millions more man hours to construct. Making something so complicated is a learning process; suppliers and builders alike will often run into unforeseen issues during production. A single misshapen part or incorrectly cut piece of fuselage can hold up a system on the assembly line, adding additional costs and causing the contractor to deliver systems late. The prime contractor (a term for the large firm who assembles and designs to product) has to juggle thousands of sub-contractors, whose parts may be of varying quality and consistency. Luckily, contractors often learn over the duration of the program, yielding savings.
Thus, large defense programs often face a multitude of hurdles during development and production. In addition, questionable procurement practices sometimes add to the risk. Prime contractors may give overly optimistic cost evaluations to win a contract, prototypes may not realistically represent the end product, governments may make unrealistic requests or fail to set deadlines, etc.
An exemplary case study in procurement overruns is the F-35 program. First of all, the F-35 was hobbled by questionable procurement practices: contractors made inaccurate price estimates, requirements were unclear, and development as well as testing schedules were unrealistic. The F-35 is also enormously sophisticated. It features millions of lines of code, many different sensors, a whole new helmet and avionics suite, stealth technologies, a massively powerful engine, a vertical takeoff fan, and various other complex components. Developing any one of these alone would be manageable, but the concurrent integration of all the components into one plane proved highly difficult. Engineers often realized that they needed far more time and investment than provided for in the development timeline, which caused delays and cost hikes. In tests, various unforeseen issues emerged, requiring costly redesigns. Supplier issues abound as well; many F-35 suppliers deliver parts late and/or below quality thresholds.
All these factors created price increases. This forced some nations to cut their orders, which boosted per-unit costs by as R&D costs were distributed between fewer planes. Now, the program is nearly complete, but the F-35 is many years behind-schedule and billions of dollars over budget.
Fortunately, there are many strategies that can reduce risk in system development. To illustrate these, the Flight III Arleigh Burke-class destroyer will be used as a case study, as its development includes many lessons learned from previous programs such as the F-35.
The easiest way to reduce risk is avoid designing new components when possible. For example, the Flight III Arleigh Burke needed new A/C units to cope with additional heat generation. Instead of designing a new A/C unit for the Burke, the engineers merely borrowed the large A/C units used in the San Antonio-class amphibious transport docks. While there are some costs associated with fitting the Antonio’s A/C units into the Burke, these costs are far lower than the cost of a clean-slate A/C design, thus saving a large amount of money and eliminating the risk associated with a new A/C design. The Flight III also uses generators and other electrical components previously found on other ship designs. Unfortunately, it is not always possible to re-use existing components, as there are often no existing designs which meet requirements for the project. Nevertheless, re-using existing components whenever possible is an effective way to reduce program costs and risk.
Another strategy increasingly embraced by the military is the usage of pre-existing commercial components in military systems. These commercial components are described as Commercial Off-The-Shelf (COTS). For a part to qualify as COTS it has to be non-custom (ie: not designed specifically for a certain application). In many fields, such as microprocessors and software, the military simply cannot match the private sector in research and development funding. Companies such as Intel and Microsoft are often on the cutting edge of their respective industries, and paying military contractors to develop custom microprocessors or operating systems would be a waste of funds. Instead, the military re-purposes readily available COTS components and software to fit its specifications. COTS parts are commonly used in applications such as computing and information technology, which is unsurprising, as private sector offerings for these products are numerous. COTS solutions are less risky than custom-made solutions because the most robust product can be chosen from a range of competitors, and the effectiveness of the product can be easily evaluated prior to any investment, reducing the risk associated with developing an untested bespoke solution.
The Arleigh-Burke class employs COTS microprocessors for radar signal processing and other functions. It also uses an open x86 computing architecture which allows for easier software upgrades down the line. The usage of these common COTS components not only saves on development costs but will likely result in lower maintenance and sustainment expenditures, as common commercial components are easier to service, upgrade, and replace than bespoke ones.
Unfortunately, for some components, such as composite armors and advanced radars, COTS is rarely an option; these components are mission-critical and are generally built to exacting military specifications. Also, there is no open commercial market for these components, so purchasing a pre-made product is not possible.
One last cost-reduction measure is the usage of high-reliability modular components. The Flight III Arleigh Burke-class destroyer uses a modular gallium nitride (GaN) AESA radar, which has a superb mean time between failures of around 100 million hours. The design is also modular, which allows small portions of the radar to be replaced without the need to service the whole antennae, saving on maintenance costs. The modular nature also allows for scalability, as future surface ship classes could use the AMDR’s modular radar units to build an antenna of any size. While higher quality solid-state components may be more expensive to develop and procure, they can often yield lifetime savings as well as performance gains, and modularity allows for flexibility in maintenance as well as future usage. For a more in-depth account of the benefits of flexibility, see a previous article on the Virginia Payload Module.
Also worth mentioning are procurement strategies and contracts. While there is much to be said about using sound procurement frameworks to reduce risk, only a few basics will be mentioned here. For one, the DoD needs to avoid its tendency to be overly optimistic. People involved in the procurement process have a disposition towards giving a rosy picture of their program in order to attain funding and priority access to resources. Unfortunately, this creates massive problems when programs underperform and drain the DoD’s resources. The usage of realistic models and more objective assessments can aid in forming more realistic program expectations. Third parties such as the Government Accountability Office (GAO) can also aid by providing a more impartial assessment of programs. Another noted strategy is rating suppliers based on past performance. This incentives realistic assessments on behalf of the supplier, as behind-schedule or over-budget deliveries hurt supplier ratings and thus future sales prospects. The Coast Guard used this strategy to choose a builder for its Offshore Patrol Cutter. Lastly, the government can choose contracts which place more of the burden on firms when overruns occur. This is a double edged sword, however, as transferring risks to the contractor can disincentive innovation.
Thus, while development of military systems is highly difficult and risky, there are measures that can be taken to reduce the likelihood of costly overruns whenever possible. As the US military re-focuses on fighting high-end conflicts, these practices will become even more critical for the development and procurement of high-end assets.
The GAO Selected Acquisitions Report, a compendium of major US military acquisitions.
GAO report on solidifying requirements earlier in the developmental process.
Congressionally mandated GAO report on the F-35, which includes many accounts of procurement issues.
Brief summary of Flight III Arleigh Burke program.