In March of 2004, just outside of Barstow, California, a series of driverless, robotic vehicles set out across the desert for Las Vegas. One after another they came to a halt, and the $1 million prize for the Grand Challenge went unclaimed. But the race isn’t over yet.
Forty self-navigating vehicles were chosen Monday by the Defense Advanced Research Projects Agency (DARPA) to compete in an October 8 sequel to last year’s first-ever autonomous on and off road vehicle challenge across the Mojave Desert.
UCLA’s team, dubbed the Golem Group (named after their partnership with the Group), successfully passed the second qualification stage.
UCLA’s team includes UCLA faculty and graduates from computer science, mechanical engineering and electrical engineering. Two of the graduate students who worked on last year’s vehicle are back for another try – Jason Meltzer and Eagle Jones. Last year’s unmanned vehicle traveled 5.2 miles, and was one of the best performers in the field.
For their second attempt, the graduate students will design and equip two vehicles simultaneously – one as a contender and one as a back up. The vehicles cannot be controlled remotely and must rely on global positioning, various sensors, lasers, radar and vision systems, or cameras, to orient themselves and detect and avoid randomly placed obstacles.
This year, the stakes are higher. The prize money has been doubled to $2 million after none of the 15 contestants in last year’s race were able to complete the difficult desert course.
The semifinalists will compete head-to-head at the California Speedway in Fontana in September and October in a series of intermediate trials designed to test the self-navigation skills of the robotic vehicles. Only half of the semifinalists chosen this week will actually qualify for the final race in October, based on how they perform, without human help, in the trials.
For the 20 teams competing in the final challenge, the first robot to make it across 150 miles of desert between Los Angeles and Las Vegas in less than 10 hours wins. The exact course is kept a highly guarded secret until just two hours before the actual race time.
This year’s group of semifinalists includes most of last year’s participants. The teams come from 16 states across the U.S. and Canada. Nearly 200 teams applied for this year’s challenge. The group of 118 accepted contenders included MIT, Princeton, the University of Washington, the University of Pennsylvania, Dartmouth, the University of Texas as well as Berkeley, Stanford, Cornell, Caltech, CMU and 5 of the UC campuses. Not all have made it past the second qualification round.
In a press release put out by DARPA, Grand Challenge Program Manager Ron Kurjanowicz praised all 118 entrants and their sponsors, and underscored that the 40 semifinalists were selected from a very strong field of competitors. “It is truly remarkable how much progress the Grand Challenge teams have made in a relatively short period of time,” he said. “The NQE will be very exciting and we will see autonomous vehicle performance that was not possible a year ago. The teams’ creative sparks are flying and they are making impressive progress toward DARPA’s goal of developing technologies that will save the lives of our men and women in uniform on the battlefield.”
To find further information about UCLA’s vision lab, go to http://www.vision.cs.ucla.edu/.
Main Image: The Golem Group’s DARPA Vehicles: from left, Golem 1 and Golem 2.