the DARPA Grand Challenge.3 The competition required autonomous ground vehicles to execute a challenging on- and off-road course from California through the desert to Nevada. The challenge offered a prize of $1 million to the vehicle that completed the course in the fastest time. DARPA funded the challenge field exercise itself and laid out funds to support the prize. No funding was provided to the participants from DARPA to support their work, yet DARPA still had more innovative participants than had been imagined and than could actually be handled during the event. While no prize winner emerged from the March 2004 challenge and no entrant completed the course, DARPA has great hopes of one day seeing its successful completion. DARPA will repeat the event in early 2005 with a new prize of $2 million.

Welby mentioned that the agency worked very hard to seek proposals from companies of all sizes and recognized that small companies were often sources of great innovation. To overcome some of the limitations under which small companies operate, DARPA participates very aggressively in the Small Business Innovative Research (SBIR) and Small Business Technology Transfer (STTR) program. DARPA program managers work very closely with small firms to grow and mature their technical and business capabilities. Welby mentioned that the agency strongly stressed the need for program managers to work with small businesses and to help integrate their efforts into the larger scope of a project’s development. DARPA's use of these mechanisms is a rich source of opportunity not only for the agency but also for participating companies.

In addition to bringing in university personnel as program managers, DARPA funds a very substantial quantity of pure and applied research at universities. Universities also participate as team members or subcontractors in large teams that execute demonstrations of large systems. In all cases, the focus is on performance. While DARPA often uses graybeard advisors and expert red teams to aid in the execution of the programs, the quality of the work is ultimately judged by the program managers, their office directors, and the head of the agency.

DARPA has a culture that depends first on technically cognizant program managers and second on performing contractors who realize that a program manager's single priority is the ruthless pursuit of the program's success. Failing programs or contractors will be cut from a DARPA program. No projects or performers are so sacrosanct that poor performance is tolerated. The flip side of this willingness to stop failing efforts is that resources continuously become available to support new and emerging opportunities.

Switching to the topic of risk management, Welby noted that DARPA seeks to identify and develop transformational technologies and systems concepts for national security. It knowingly invests in high-risk and high-payoff projects. That means that DARPA executes projects that cover a very large range of maturity and risk, from basic research in mathematics to biology to construction of prototype aircraft.

Because DARPA generally transitions products to external development acquisition organizations that specialize in final product development, it tends to aim for managing risk rather than achieving maturity. Programs are structured to remove key technology risks in a concept or product rather than to demonstrate a final refined product. DARPA uses a variety of mechanisms and methods for identifying risks, but Welby said it had no specific, formal, agency-wide standards. Program managers often work with their


Further information on DARPA’s Grand Challenge may be found at <>. Accessed September 27, 2004.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement