“Urban Challenge” is a competition that requires autonomous vehicles to navigate 100 kilometers in a mock urban environment without a driver. The team whose car manages to complete the race fastest and with no complications will receive $2 million. The autonomous vehicles must be able to navigate in traffic, reacting appropriately to e.g., stop signs, intersections, and closed streets. The vehicles must be able to recognize all of these traffic situations while following the prescribed route, and their sensors and computers must react “intelligently,” i.e., as a human driver would.
The Urban Challenge is the third major competition in the Grand Challenge series for autonomous vehicles. In 2005 the main competition was to navigate 200 kilometers in the desert. Each year the level of difficulty is increased. The race in California may be viewed in the Internet.
The competition is a real marathon: In 2006, 87 teams registered, mainly from the U.S., but also some from Europe and Asia. In April this year, the number of teams was reduced to 53 on the basis of videos of the autonomous vehicles in action. In June and July, the cars were inspected during site visits, and the number of participants was reduced to 36 semifinalists. During the site inspections, the vehicles were observed in action for a period of four hours.
“Spirit of Berlin,” Freie Universität Berlin’s autonomous car was tested on site on June 26 in San Antonio, Texas. For this purpose the car had to be flown from Berlin to Houston because there was not enough time for shipping it. Freie Universität’s team has been in Texas since June and is busy preparing for the next step of the competition.
Freie Universität Berlin’s autonomous car project is a joint effort with the Berlin Police Department. The Berlin Police Department expressed the need for robotic vehicles for the surveillance of larger enclosed areas such as warehouses or even airports. In May, 2007, the car was presented to the public. In the short term the technology developed for such projects can be used for driver assistance systems that warn drivers of dangers on the street or freeway, or to optimize the vehicle’s fuel consumption. In the long term, autonomous vehicles will be safer than those driven by humans.