Computing Research Policy Blog

CS Profs and the DOD

Long-time readers of this blog, or anyone familiar with CRA's policy efforts, will know that we've spent a lot of time raising concerns about policy shifts at the Defense Advanced Research Projects Agency (DARPA) that have cut university participation rates in DARPA-funded computer science research. In congressional testimony and blog posts, we've pointed out that a shift at DARPA -- a focus on nearer-term efforts with an emphasis on go/no-go milestones at relatively short intervals and an increased use of classification -- has sharply reduced the amount of DARPA-supported research being performed in U.S. universities. In fact, between FY 2001 and FY 2004 (the last year for which we have good data), the amount of funding from DARPA to U.S. universities for computer science research fell by half -- and informal evidence suggests university shares are even lower today.

There are a number of reasons we're concerned about this trend. For one, DARPA's diminished support for university CS leaves a hole in the federal IT R&D portfolio -- both in funding, but maybe more importantly, in the loss of the "DARPA model" of research support. Since the early 1960s, the country (indeed, the world) has reaped the benefits of the diverse approaches to funding IT research represented by the two leading agencies -- NSF and DARPA. While NSF has primarily focused on small grants for individual researchers at a wide range of institutions -- and support for computing infrastructure at America's universities -- DARPA's approach has been to identify key problems of interest to the agency and then assemble and nurture communities of researchers to address them. The combination of models has been enormously beneficial -- DARPA-supported research in computing over the last four decades has laid down the foundations for the modern microprocessor, the internet, the graphical user interface, single-user workstations and a whole host of other innovations that have made the U.S. military the best in the world, driven the new economy, changed the conduct of science and enabled whole new scientific disciplines.

But DARPA's policy shift also impacts its own mission, which is to ensure the U.S. never again suffers the sort of technological surprise marked by the Soviet launch of Sputnik (which motivated the establishment of the agency nearly 50 years ago). DARPA's move away from support of university researchers means that many of the brightest minds of the country (indeed, the world) are no longer working on defense-related problems. This loss of mindshare -- the percentage of people working on DARPA-related problems -- is very worrisome to those in the community who understand how much of America's advantage on the battlefield (and in the marketplace) is owed to a network-centric strategy. I hear concerns from the "old guard" in many of America's top university CS departments that there's a whole generation of young researchers who have no experience working with DARPA or the Defense Department and who are not attuned to defense problems -- a fact that doesn't bode well for the future of the U.S. technological advantage and DARPA's goal of preventing technological surprise.

To their credit, the folks at DARPA recognize that this lack of awareness among younger faculty of the types of problems DARPA would really like to solve is a situation that needs addressing. And one way they're approaching the problem is very direct -- they're finding young faculty with research areas of interest to the agency and, well, taking them on a little tour of the DOD. The Computer Science Study Group, run by the Institute for Defense Analysis for DARPA, serves to "acclimate a generation of researchers to the needs and priorities of the DOD," by mentoring, holding workshops, field trips to DOD facilities and fairly elaborate (and pretty kewl) show-and-tells. An interesting article today on Rensselaer ECSE professor Rich Radke's experience has some details on CSSG goals and methods:

The multi-year program familiarizes up-and-coming faculty from American universities with DoD practices, challenges, and risks. Participants are encouraged to view their own research through this new perspective, and then to explore and develop technologies that have the potential to transition innovative and revolutionary computer science and technology advances to the government.

"The basic idea is to expose young faculty to Department of Defense-related activities, via briefings by military and intelligence officers and ‘field trips’ to military and industrial bases," Radke said. "It is truly a hard-core experience filled with days of interesting briefings and up-close show-and-tell with vehicles and equipment."

2007 was the first year for the CSSG and the $4.5 million program supported about a dozen young researchers. DARPA has requested an increase in the program for FY 08 ($7 million) and FY 09 ($7.7 million), so hopefully we'll see that number start to rise.

The DARPA CSSG program is one part of addressing the overall problem. The larger concern is the importance of bringing DARPA back into the university research fold -- not because it would benefit academic researchers, but because it impacts the mission success of the Department of Defense (and hence our national security). A number of factors suggest that maybe it's time to focus on the goal of increasing mindshare of the best brains working on U.S. defense-related problems. For one, because of U.S. visa policies, increasingly the best minds in the world won't necessarily be coming to the U.S. Second, the research capacity of our potential adversaries increases daily. And finally, the increase in foreign investment in U.S. university research departments means that competition for U.S. university mindshare is only increasing, and in some cases, maybe from countries we'd rather not gain a competitive leg-up on us. So, programs like CSSG are really important. But maybe so are some bigger policy issues across the agency....