The most recent Urban Challenge was held Nov. 3, 2007, and consisted of unmanned cars racing around an old Air Force base. Eleven unmanned cars were in the race, and 30 cars with drivers were added to the course to increase the traffic density. Teams had no prior knowledge of the race course. They were given a map of the course two days before, but this wasn't a simple point-A-to-B race. The vehicles had to visit checkpoints in a specific order, like an unmanned military vehicle would have to do as a part of a mission. DARPA officials were on hand throughout the course to observe the vehicles and monitor safety.
The cars used a number of computer programs, as well as radar sensors and Global Positioning Systems (GPS). The systems all worked together to calculate a course, sense obstacles and avoid other cars. Some systems are similar to technology already in cars on the road, like Automatic Cruise Control, which keeps a set distance between a car and the car in front of it. Other technologies, which allow a car to come to an intersection, sense other traffic and know who has the right of way, are years from mass production.
Though 11 vehicles started the race, by midmorning, half had been disqualified. One was stopped moments before it ran into a building, and another simply pulled off the road and into a carport before being disqualified. In the end, six vehicles finished the race: Tartan Racing (a team put together by Carnegie Mellon University and General Motors, among others), won first prize (and $2 million), while teams from Stanford University and Virginia Tech came in second and third, respectively.
All in all, the 2007 DARPA Urban Challenge showed that technologies for unmanned vehicles are closer than you might think. While there are a lot of bugs to work out of the technology, the same can be said for human drivers.
To learn more about car technologies, look over the links on the next page.