PITTSBURGH—Carnegie Mellon University will partner with General Motors Corp. to form Tartan Racing, a team that will enter a driverless Chevy Tahoe in the $2 million Defense Advanced Research Projects Agency (DARPA) Urban Challenge scheduled for November 2007.
The race will require autonomous vehicles to travel 60 miles of streets in a mock urban setting. To succeed, vehicles must drive completely on their own — without drivers or remote control — and finish the course within six hours. Cars must stay in their lanes, negotiate intersections, drive in traffic and follow rules of the road with only computers at the wheel.
The Urban Challenge is designed to develop autonomous driving skills for city streets, eventually leading to cars that drive themselves. In the short term, the experimental technology developed for the challenge could yield new devices that aid human drivers and improve highway safety.
"The Urban Challenge will develop a leap of capability beyond what is possible in today's human-driven cars," said William "Red" Whittaker, the Fredkin Research Professor at Carnegie Mellon's Robotics Institute and team leader for Tartan Racing.
Since there are no human drivers in the Urban Challenge, the driverless cars must "see" roads and other vehicles with cameras, lasers, radar, sensors and other breaking technologies. Planning software continuously determines where and how to drive, how to stay out of trouble and how to quickly reach a destination.
Tartan Racing's Chevy Tahoe will be equipped with automated throttle, brakes and steering for computer control of physical motion.
"A challenge like this requires not only the best technologists, but also the best sponsors," Whittaker said. "With General Motors as our premier sponsor, Tartan Racing draws on the support and unmatched technical expertise of the world's largest automotive company."
"The rapid progression from stand-alone safety and convenience features to integrated driver-assistance systems will soon make autonomous driving a reality," said Larry Burns, GM vice president for R&D and Strategic Planning. "Autonomous driving has the potential to dramatically enhance vehicle safety, customer convenience and fuel economy, and is a key component of GM's vision to reinvent the automobile using advanced technology.
"We are very excited to be partnering with Carnegie Mellon on the Urban Challenge. Our partnership will accelerate our capability in this fast-moving technology area and the challenge will be an important testing ground for our autonomous driving systems and enabling technologies," Burns said.
Chris Urmson, Tartan Racing's technology leader, will direct a team of world-class researchers drawn from Carnegie Mellon and GM. These include Carnegie Mellon Robotics Institute (RI) Research Professor Anthony Stentz; RI Professor Martial Hebert; RI Senior Systems Scientist John Dolan; and Raj Rajkumar, Carnegie Mellon professor of electrical and computer engineering and computer science, who is co-director of the GM-Carnegie Mellon Collaborative Research Laboratory.
The team also includes Bakhtiar Litkouhi, acting director of the Electrical and Controls Integration Lab at GM R&D and co-director of the GM-Carnegie Mellon Collaborative Research Lab; Varsha Sadekar, group leader, active safety and driver assistance, and program manager for the Urban Challenge; and other researchers and technologists from General Motors.
Whittaker said Tartan Racing will continue to seek additional sponsors and teammates as it prepares for the event, now scheduled for Nov. 3, 2007.
DARPA says its Urban Challenge will be at an as-yet undisclosed site somewhere in the western United States. It will award a $2 million prize to the team whose vehicle completes the course fastest in less than six hours, with $500,000 and $250,000 prizes going to the second- and third-place finishers, respectively.
In 2004 and 2005, Carnegie Mellon teams led by Whittaker competed in DARPA Grand Challenges, in which robotic vehicles followed desert courses.
Byron Spice | 412-268-9068 | bspice [atsymbol] cs.cmu.edu