Innovative Management in the DARPA High Altitude Endurance Unmanned Aerial Vehicle Program: Phase II Experience
Jan 1, 1999
Lessons Learned from the HAE UAV Program
Historically, U.S. unmanned air vehicle (UAV) development programs have been bedeviled by cost growth, schedule slippage, manifold technical deficiencies during protracted development, and generally disappointing operational results.
Despite these problems, high-level support for UAVs persists; thus, there is a strong motivation to find ways to overcome these programmatic and technical difficulties. As one potential solution for these historical problems, the Defense Advanced Research Projects Agency (DARPA), with congressional support, initiated an innovative acquisition program that differed significantly from the normal Department of Defense (DoD) procedure. The program, which DARPA transferred to the Air Force in October 1998, centers around developing two UAVs within the High Altitude Endurance (HAE) UAV program office—a conventional UAV known as "Global Hawk" and a low-observable UAV known as "DarkStar."
This research has focused on understanding how the acquisition innovations have affected program outcomes so far and on identifying lessons learned that might be applied to improve DoD acquisition strategies.
The acquisition strategy for the HAE UAV program has four key innovative elements:
While other programs have included ACTD or IPPD experience, the HAE UAV is unique in combining all these elements into an acquisition strategy.
Assessing the effect of these innovative elements on HAE UAV program outcomes is a challenging task. Each element has advantages (benefits) and disadvantages (costs), as summarized in the table below.
|Acquisition Elements||Key Advantages||Key Disadvantages|
|ACTD Designation||Increased design and decision flexibility||Accelerated program structure, leading to technical problems|
|Section 845 OTA||Lower cost of systems||Inadequate engineering discipline|
|IPPD Process||Reduced oversight and open interactions||Potential accountability conflict between IPPD process and Section 845 OTA|
|UFP/Performance Goals||Contractor control over cost-performance trade space; improved cost discipline||Ambiguous definition of military utility, leading to minimal use of available trade space|
ACTD Designation. The designation increased design and decision flexibility; however, bounding total program length led to reducing the user-demonstration phase by half, severely limiting the scope of user activities and the range of conditions under which the Global Hawk and DarkStar systems could be evaluated. ACTD designation provides the opportunity for early user involvement; however, it was not adequately used here. Users were informed of program status and direction, but they provided little input.
Section 845 OTA. Overall, managing the program under Section 845 OTA had a substantial effect on lowering the costs for the basic design and development efforts for the two systems. The use of contractor-based management processes (reporting, auditing, accounting, etc.) reduced overhead costs by eliminating levels of review and led to decreased decisionmaking and increased flexibility. However, the increased flexibility came with a cost—a lack of engineering discipline in the systems engineering approach, indicated by the concurrent development of systems and their specifications, software development problems, and inadequate attention to systems integration.
IPPD Process. Though somewhat strained at first, the government–industry relationship evolved into an interaction that both Joint Program Office (JPO) staff and contractors saw as an improvement over the traditional adversarial approach. The IPT structure allowed the government to obtain timely insight into program status, problems, and solutions, as well as provide timely input into the contractors' decision processes. However, there is a potential conflict between the IPPD process, which encourages teamwork and joint accountability, and Section 845 OTA, which encourages contractor responsibility and decision authority; it is not always apparent who owns the process and who is accountable for decisions. An optimum balance between oversight and technical participation has not been reached.
UFP/Performance Goals. The UFP approach improved overall cost discipline of the program; contractors theoretically had complete control over the cost–performance trade space. However, in practice, such control was severely undermined because military utility was inadequately defined up front; as a result, both the contractor and the government were reluctant to drop functionality as a way of maintaining UFP, which crippled the ability and willingness of the program to make cost–performance trade-offs.
The relationship between UFP and nonrecurring engineering (NRE) funds is another issue. To achieve a given UFP, there must be a certain level of NRE activity and investment. The balance between NRE funding and the UFP must be clearly rationalized for the UFP to be credible and achievable. To mitigate an apparent NRE funding shortfall, the contractors "created" additional NRE funding by trading off developmental activity content (which also minimizes developmental cost growth) as opposed to functionality. While both trade-offs were allowable, the government should have clarified which one it preferred and prohibited the other, because the two have very different implications for system development and technical maturity.
While not applicable to all acquisition programs, the strategy used in the program is an effective alternative to traditional acquisition processes, one that can be enhanced in several ways. First, a mechanism for government intervention to deal with contractor weaknesses—such as a lack of software development expertise—should be incorporated into future government-industry relationships of this type. Second, to overcome the cultural impediment to early user involvement, the mechanisms for such involvement should be more formalized. Third, program risks need to be clearly identified early on and a plan developed to manage them. Finally, future programs should go beyond the concept of UFP as a single goal; once bounded, cost, schedule, and performance should all be stated goals, though the resulting trade space must be large enough to allow credible trade-offs.