Climb into the passenger seat of your car, relax, and let your car drive you where you want to go. Forget steering wheel, brakes and gas pedal. Your car’s computer vision, laser range-finder and GPS systems will get you safely to your destination.
A futuristic scenario? Not as much as you might think, says John Spletzer, assistant professor of computer science and engineering at Lehigh.
Spletzer and students at work in the lab
Spletzer is working with engineers from the University of Pennsylvania and Lockheed to design and build a robotic car that can navigate roads and obstacles with no driver or remote control.
The Penn-Lehigh-Lockheed Ben Franklin Racing Team
is one of more than 80 around the U.S. that have answered a call from the Defense Advanced Research Projects Agency (DARPA), the research arm of the Pentagon.
After a series of qualifying events, the surviving 10 to 20 teams will enter their cars at DARPA’s 2007 Urban Challenge, the third event of its kind, which will be held in November at an undisclosed location.
The goal at the Challenge is to be the first car to complete a 60-mile course through a simulated urban environment strewn with obstacles that are both stationary and moving. The winning team will receive $2 million. Second place is worth $1 million and third place $500,000.
The overall goal is to help DARPA develop sophisticated, driverless, ground-combat vehicles for the U.S. military and thus meet a congressional mandate that one-third of such vehicles be unmanned by 2015.
The Penn-Lehigh-Lockheed racing team has received additional funding from Thales Communications, whose president and CEO is Mitch Herbets ’79. Thales is part of the Thales Group, a Fortune 500 company and the ninth largest defense contractor in the world.
Making a car drive like a human
For a car to navigate roads by itself, says Spletzer, it must be equipped with the same intelligence and sensing skills that human beings possess. The car must be able to recognize the lanes, median and shoulder of the road it is traveling. It must also be able to detect approaching vehicles and to distinguish between these vehicles and other obstacles.
The car’s “road-segmentation” and “lane-tracking” abilities must be particularly robust, says Spletzer. In addition to identifying and adhering to that portion of the terrain that is drivable road surface, the car must also be able to determine what parts of a parking lot it can and cannot drive through. If no lines are painted on a road, or if the car’s GPS system fails – a likelihood under bridges and overpasses and in skyscraper-dominated “urban canyons” – the car must be able to continue driving on the paved portion of the road.
“The car must at all times know where the road is, where the car’s half of the road is, and where the edge of the road is,” says Spletzer, who in a separate project is developing a self-stowing robotic wheelchair for disabled drivers.
“The car not only needs to stay in its lane and remain the proper distance behind the car in front of it, it also needs to know to stop behind double-parked cars in its lane and wait for traffic ahead to clear before it proceeds.”
To accomplish these tasks reliably, robotic cars require state-of-the-art laser and computer vision systems, two of the areas in which Spletzer is contributing his expertise to the Penn-Lehigh-Lockheed Ben Franklin Racing Team.
The team’s car, a revamped Toyota Prius nicknamed “Little Ben,” is equipped with video-camera “eyes” and laser range-finder systems called lidar, an acronym for light detection and ranging. Lidar can be used to estimate the distance and speed of clearly defined remote targets.
“We’re developing the software for obstacle detection, road segmentation and lane tracking,” says Spletzer. “This requires robust, real-time algorithms that enable the car’s systems, especially its camera eyes, to operate as reliably as a human being 100 percent of the time.”
The Penn-Lehigh-Lockheed Ben Franklin Racing Team has recently submitted a video
to DARPA that demonstrates the vehicle in operation. If the agency likes what it sees, it will make a site visit to Lehigh’s Goodman Campus over the summer to verify Little Ben’s autonomous driving capabilities. The next step is the national qualifying round in October, followed by the Grand Challenge in November.
Teams that are chosen to compete in the qualifying round and Grand Challenge, says Spletzer, stand a good chance of receiving subsequent funding from DARPA to help the agency develop robotic ground-combat vehicles for the military by 2015.
Spletzer, meanwhile, is confident engineers can go further. Just as computers have been programmed and trained to beat the world’s grand chessmasters, he says, robotic cars may someday navigate roads more ably than humans can.
“Computers can really do well when you focus them on specific tasks,” he says, citing IBM’s “Deep Blue,” which in the 1990s became the first machine to win a chess tournament against a human.
“Ultimately, we’d like to make robotic cars that can drive better than humans can. That won’t be easy. The driving environment has many complicated aspects. It’s not limited to an eight-by-eight chessboard.”
To learn more about the Ben Franklin Racing Team, contact John Spletzer
or Lehigh’s Corporate and Foundation Relations Office at 610-758-5352.
Posted on Thursday, May 10, 2007