The Defense Advanced Research Projects Agency (DARPA) hosted the first Grand Challenge Project last year, offering a reward of $1 million. This year, the prize money has been doubled, making the competition all the more interesting. We looked at how two teams are working towards the challenge, one in California called Team Cyberrider and one in Germany. The German team is composed of a partnership between Volkswagen and Stanford University. Stanford Professor Sebastian Thrun, an internationally accredited expert in artificial intelligence, has assembled an extremely competent development and research team spanning nine different time zones.

The Hummer H1 that entered last year's challenged used slightly older technology.

The Grand Challenge pits autonomous vehicles against each other in a race across the Southern California desert. With no driver, these robotic cars use the latest technology in following GPS waypoints and avoiding obstacles. Last year, none of the vehicles finished the race, which proved that the Grand Challenge really was "grand". The winding 150 mile course from Barstow, California to Primm, Nevada knocked out vehicle after vehicle. Most of the vehicles were eliminated just miles of the starting gate, with Carnegie Mellon's Sandstorm reaching the farthest distance of 7.4 miles, before clipping a fencepost and then slamming into a three-foot boulder.

While the primary guidance is supposed to be via GPS waypoints, vehicles must be able to navigate in absence of any GPS signal. Tunnels and nearby buildings can interfere and completely block signals. In addition, tank traps, large ditches and boulders may be placed along the route between two waypoints. Vehicles must decide how to navigate and avoid these obstacles while traveling at 10 to 30 miles per hour.

"This is the first long-distance race in the history of the automobile in which the vehicles have to make all of the necessary decisions about their progress themselves," said Professor Sebastian Thrun, head of the Stanford Racing Team. "In other words, the car doesn't just need a strong body, but an especially smart head as well."