Humans won't be announcing until today how the robots finished in the Urban Challenge race in California, but Carnegie Mellon University is hopeful that its driverless Chevrolet Tahoe -- nicknamed "Boss" -- earned first place and the $2 million that comes with it.

CMU's Tartan Racing team, billed as a race favorite, crossed the finish line one minute behind Stanford University's team. CMU started about 20 minutes after Stanford in the competition through a cityscape on a former Air Force base in Victorville, Calif., with two-way traffic, stop signs, four-way intersections and other roadway obstacles.

However, the 11 robotic vehicles are not only being judged on finishing the six-hour course but in making the most efficient run with the fewest mistakes or traffic violations -- the California driving code was in effect. The race sponsor, the Defense Advanced Research Project Agency (DARPA), has to calculate all of that to determine the exact order of finish.

Chris Urmson, CMU's Tartan Racing director of technology, was ecstatic in a telephone call after the competition yesterday.

"Boss did really well. It was driving itself safely and fast. It looked like a race car. We're very excited," Mr. Urmson said. "It's very possible we came in first but at this point we're very thrilled because it was a very challenging race and it was able to drive itself around and come back.

"It's a huge day for the field of robotics, There were a lot of robots running out there safely, 11 at one point, and for the most part you couldn't tell they didn't have a driver."

The Urban Challenge is the third robotic vehicle competition that DARPA has held since 2004 to develop technology to save lives on the battlefield. In addition to the $2-million first prize, second place earns $1 million and third place gets $500,000.

To help it drive safely through the mock city, Boss has 18 sensors and 300,000 lines of new computer software that make driving decisions 20 times a second.

Boss travels at 30 mph, or 13 yards per second, leaving little margin for error. But it reasons what other vehicles are doing, where it's situated, where it needs to go and the best route to get there.