Kelman: The grand challenge
DARPA offered $2 million for the first driverless vehicle to cross the finish line
- By Steve Kelman
- Nov 21, 2005
There was a good procurement news story recently about the second running of the Defense Advanced Research Projects Agency's Grand Challenge. The Grand Challenge is a race in which entrants -- 23 of them this year -- develop driverless vehicles designed to navigate a 140-mile course through the desert.
DARPA offered a $2 million prize to the first vehicle to make it to the finish line. Seeking to encourage development of technology that can be used for riderless vehicles in battle, the agency held the first Grand Challenge in March 2004. That time, no vehicle was able to finish. This time, five did. The contest stimulated technological progress.
The Grand Challenge could be seen as an innovative procurement process -- procurement by contest. Indeed, at the time of the first one, I wrote about it in a Federal Computer Week column.
However, there's been a lot of water under the procurement bridge in the past 20 months. Since the first Grand Challenge, the attention of politicians and reporters has been on other things, such as real and alleged procurement scandals.
The nation is lucky that the Grand Challenge developed before the recent political and media focus. Would the Grand Challenge have much chance of getting off the ground today? I don't think so.
For starters, in today's environment, people are being encouraged or intimidated to put their time and energy into keeping their noses clean to avoid doing something wrong, instead of thinking about better ways of doing business so they can do something right.
Beyond that, today the Grand Challenge would be seen as risky. The biggest problem would be that DARPA was offering the winner $2 million without regard to how much the individual or team had spent.
What if the winner had spent only $50,000 plus graduate student volunteer time, thus making a profit of something like 4,000 percent? (I don't believe this was the case, but the government had no way of knowing in advance.)
Contest rules didn't allow government auditors to check the winner's costs. Think of the risks the agency runs of being targeted by a fleecing-of-America story or a press release from a watchdog group.
Furthermore, the government did not prescribe and inspect for use of specific designs or methods. In other words, this was a performance-based procurement, described recently in the Washington Post as a nontraditional approach -- heaven forbid! -- requiring extra vigilance so government isn't bamboozled.
Simply as a new buying method, the Grand Challenge created risks DARPA might not have known about initially, but they could come back to haunt it.
Today the event would be unlikely to be conceived in the first place, and if it were proposed, it almost certainly would not be approved. People would spend their energy on other aspects of procurement. Is this how we want things to be?
Kelman is a professor of public management at Harvard University's Kennedy School of Government and former administrator
of the Office of Federal Procurement Policy. He can be reached at [email protected].