Carnegie Mellon wins Urban Challenge

Carnegie Mellon University’s Tartan Racing Team has won the DARPA’s 2007 Urban Challenge competition — netting the $2 million first place prize. Stanford University took second place and $1 million, and Virginia Tech took $500,000 for third. The Register’s Ashlee Vance reports that Six vehicles of the eleven vehicles finished the sixty-mile course with Stanford, Carnegie Mellon, and Virginia Tech all closing within a few minutes of each other. The robots had to complete a number of different missions, returning to the starting point each time to be given a new task. Teams from Ben Franklin Racing Team, MIT, and Cornell finished as well, although after the alloted six-hour limit. The contest proved a success for DARPA and for the advancement of robotics technology. Only three short years ago, robotic vehicles failed to travel more than a few miles along a desert course. In that race, the systems were hand-fed GPS points and needed only to focus on speed while avoiding stationary targets such as ditches and trees. One year later, the vehicles mastered those tasks but were unprepared for facing more demanding tasks. Stanford won the second race, beating out two vehicles produced by longtime robotics powerhouse Carnegie Mellon. Last year’s Stanford victory was especially poignant for Carnegie Mellon because the Stanford team is led by Sebastian Thrun, a former student of Carnegie Mellon’s team leader Red Whittaker, a professor of robotics at Carnegie Mellon and an innovative pioneer in the field. This year’s victory by the Tartans evens the score. In addition to the money, the winning teams received a rather ugly giant eagle statue. The statue weighs about 100 pounds, so winners have great difficulty, to say nothing of the risk involved, raising it over their heads

DARPA officials said that Carnegie Mellon averaged about fourteen mph with Stanford and Virginia Tech slightly under that. During the past two years, university-based robotic race teams have shown the most progress. DARPA’s first event two years ago — the Grand Challenge — was criticized for allowing too many teams, some with but scant technical qualifications, to participate. Now that DARPA has done a much better job in qualifying the participants, and with teams proving that they can learn from their mistakes, the idea of robotic driving machines no longer appears to be the stuff of fantasy as it did two years ago. DARPA used various criteria to determine the winner, looking at how quickly the robots completed the course and how many traffic violations they endured. The organization collected this data from the vehicles and by watching the robots from the sky with an airplane. That plane sent back video with up to .5m resolution of the event, allowing officials to replay incidents in slow motion as needed to determine who was at fault.

It is not clear that there will be another Grand Challenge, as the government has taken away DARPA’s authority to offer cash prizes for these types of events without prior approval. “I don’t know why they did it, but they did it,” DARPA Director Anthony Tether said. “I don’t have the authority to say, ‘Yes’. We never really finish anything (at DARPA). We just show that it can be done. We take the excuse off the table. I think we are close to that point,” he said. Red Whittaker has called for a twenty-four-hour race which would force vehicles to deal with different terrains, weather, and day and night conditions. According to Stanford’s Thrun, these events prove crucial to U.S. technology development. They help encourage youngsters’ interest in robotics. “We are going to build a whole new student force for the next thirty years to come.”