CONTACT: Stanford University News Service (650) 723-2558
STANFORD -- By taking an aging robot and stuffing its circuits full of artificial intelligence, a team of Stanford computer science students managed to capture first place in one of two events at the National Mobile Robot Contest in Seattle.
The annual contest, sponsored by the American Association for Artificial Intelligence, was held in Seattle from July 30 to Aug. 2. It had two events: trash pickup, in which the robots had to pick up all the trash in a given area and put it in a trash can, and office delivery, which required the robots to navigate between rooms in an office environment while avoiding furniture and other realistic obstacles.
Robots from 15 different institutions - including Carnegie Mellon University, the University of Bonn, and Lockheed Missiles and Space Division - competed. The Stanford programming team consisted of undergraduate Stan Birchfield, newly graduated Rob Powers and doctoral student Illah Nourbakhsh from the logic group in computer science.
Dervish, the R2D2-sized Stanford entry, was the only contestant from the field of 11 robots to successfully complete the office delivery task in the contest finals. Each robot was given a schematic map of the office layout without dimensions or other geometric information. The simulated office environment consisted of 10 cubicle offices and six hallways. Contestants were taken to one of the offices and disoriented (the robot equivalent of spinning someone while she is blindfolded). Then each robot was expected to make its way to another one of the offices as quickly as possible, avoiding obstacles and identifying offices, whether the door was open or closed.
The Stanford robot, which not only was old but was equipped only with sonar, the cheapest type of robot sensor, was not very fast, Nourbakhsh admits, but it had enough "smarts" to finish the job. Other robot-contestants were newer, faster and had more sophisticated sensor arrays, but the problem was too difficult for them. As it moved from office to office, Dervish also sang an eight-stanza robot song. Sorry, it wasn't the robot classic "Daisy."
This is the second year that the Stanford entry has walked away with one of the blue ribbons at this competition. "The first time many people thought it was a fluke. Now we've shown that it wasn't," says Nourbakhsh.
This is an archived release.
This release is not available in any other form.
Images mentioned in this release are not available online.
Stanford News Service has an extensive library of images, some of which may be available to you online. Direct your request by EMail to firstname.lastname@example.org.