The Wright brothers must have been smiling on Oct. 8 when teams of inventors set out in the Nevada desert to prove to the world that a car could travel over 131 miles over a varied and harsh terrain, not only minus a driver, but without any human input.

The five teams who completed in the Defense Advanced Research Projects Agency race traveled to the Pentagon courtyard this week to show off their creations. Stanley, a Volkswagen Toureg representing Stanford University, finished first in the DARPA Grand Challenge and earned a check for $2 million.

Last year DARPA put up $1 million in incentive money, but no team was able to go more than seven miles. Spectators cheered for the competitors and the race garnered wide publicity, but it was all for a very serious reason.

"We established the Grand Challenge program to help foster the development of autonomous vehicle technology that will some day help save the lives of Americans who are protecting our country on the battlefield," said DARPA Grand Challenge program manager Ron Kurjanowicz on their website. "The outcome of this great public event demonstrates that we have succeeded in our mission."

Gabe Hoffmann, 26, of the winning Stanford team, said his group learned a great deal from last year's race.

"There is so much competition, I was surprised we won, not that we finished," he said. "The biggest hurdle was software."

Hoffman explained the VW's intelligence consisted of three different levels of sensing devices. The first is LIDAR, or laser sensors, to scan the road and create a terrain map. The second level of sensors is a camera and radar to test the road ahead and tell the computer to slow down or speed up.

The third sensing component consists of an inertial measurement unit to sense the turning and acceleration rates and a GPS system for position and orientation.

The information from the sensors goes into a computer in the back of the car, which then determines where to go, the speed and to brake when necessary.

"Testing is critical," Hoffmann said. "We spent months in the desert. If someone walked across in front of it, it would remember and always avoid. This thing has a mind of its own."

Josh Struble, an engineer for the Caterpillar Corporation, of the Carnegie Mellon Red Team second -- a Hummer -- and third place vehicles -- a Humvee -- explained some of the problems and possible uses for the unmanned vehicles.

"It hits things a human would drive around," he said. "It's not the smartest driver. They could be used for repetitive tasks that humans do. They would do supply routes, perimeter patrol, border patrol. DARPA wants one third of the military's ground vehicles unmanned by 2010. It's a pretty bold goal."

Not all of the finishers were backed by large research universities and big corporate sponsor. The Gray Team's Kat-5 finished fourth. They consisted of a group of IT guys from a small insurance company in Metairie, La., and some students from Tulane University with no experience in robotics or artificial intelligence.

Powell Kinney, an undergraduate at Tulane in bio-medical engineering, claimed a software bug prevented them from winning.

"Because of Katrina, we lost testing time," he said. "Had we not had the bug we would have beaten Stanford. The consensus is we would have. "

The Gray team spent $650,000 to outfit their Ford Escape hybrid vehicle for the challenge in six months. While other well-funded teams tested their vehicles in the desert for months, Kat-5 drove around Hammond, La., Most of the team lost homes due to Hurricane Katrina and their base of operations was an old Coca Cola plant.

"Our approach had to be limited, simple solutions," Kinney said. "We lost three weeks and a lot of equipment because of Katrina. We're very proud of Kat-5."