Robots Compete for $2 Million DARPA Prize

Tires spun, bales of hay caught fire, and robots crashed, but 23 unmanned ground vehicles have qualified for the Defense Advanced Research Agency's Grand Challenge this weekend.

The robots, built by teams from New York to California, will race across 150-miles on an undisclosed route, traversing dry lakebeds, speeding over steep mountain terrain, and ambling around obstacles on the desert route. The route is being kept secret until two hours before the race, expected to begin at 6:30 a.m., PDT, Saturday, near Primm, Nev.

The professors, students, engineers and other experts who designed the robots aren't just competing for the pride that victory bestows.

DARPA, the agency that launched the federal government's effort to build the Internet, is offering a $2 million prize to the group whose robot is the first to complete the route in less than 10 hours.

But that's not why William "Red" Whittaker, named by Science Digest as one of the country's Top 100 U.S. Innovators, is doing it. Once named Pittsburgh's Man of the Year in Technology, he holds 16 patents and has written or co-authored more than 200 publications. He has been out West all week, leading competitors from Carnegie Mellon University.

"I was born to do this," he said in a phone interview Thursday. "I love it. I am enchanted by the idea of machines that sense and think and act to work in the world -- machines that farm, machines that explore, machines that mine. It's like flying to the moon, or flying to Paris for the first time. No one has done it yet, so no so it's not clear what it takes to get the job done. It's not just about one race or one date."

DARPA Director Tony Tether said the technology behind the robots could someday save lives. The military wants one-third of its vehicles to be driverless by 2015. The research behind the Minivans, Volkswagons, Fords and light strike vehicles bobbing, weaving, and turning over obstacles and around sharp bends could be put to use on battlefields.

Cameras, radar and sensors "see" the route and sense obstacles, customized software, and GPS with dead reckoning, guide the machines over the rugged terrain, through tunnels, past garbage cans and around hay bales in their path.

Source