Robotic vehicles competing in the DARPA Grand Challenge fell far short of completing the race, going just a few miles at best. But this year’s contestants feel their unmanned vehicles may be able to complete the 150 mile course through the desert.
“There’s a very real possibility someone will win this year given that we’ve had another 14 months,” says Brett Leedy, captain of the Virginia Tech team. “Last year’s teams didn’t have enough time. We didn’t even start testing our vehicle until a few weeks before the race.”
He is among the graduate students from Virginia Tech demonstrating Rocky, one of the university’s two vehicles that recently made the cut down to 40 robotic racers, at NIWeek. Nearly 100 entries have already been cut by DARPA, and only 20 autonomous vehicles will actually get to run the race on Oct. 8.
Though he’s hoping Rocky or its close relative, Cliff, will win, he’s far short of predicting a victory. But he notes that last year's teams barely had a year to accomplish everything, while this time they have had more than a year to push their designs forward and test the results.
Va. Tech just completed a rigorous test in New Mexico, analyzing the performance of new and existing equipment. Rocky now incorporates an additional laser scanner, giving it three ways to scan the terrain. A stereo vision systems’ two lenses will also help rocky avoid the boulders it’s named after.
To help the vehicle chart its way along a route that won’t be divulged until two hours before race time, Rocky’s GPS and inertial guidance sensors have also been upgraded, according to Michael Fleming.
To help make sure that all these electronics operate in the harsh desert environment, the team of about three dozen undergraduates and a handful of grad students are also improving the filtration system on the electronic box that holds four PXI-based systems. Additional cooling techniques are also being used.