November 9, 2007 By Chad Vander Veen
Photo: Carnegie Mellon's winning robot "Boss"
Seventy miles from Los Angeles, in the Mojave Desert, a race unlike any other was held. For two weeks in late October and early November, teams from around the country competed for a $2 million purse in the third DARPA Grand Challenge.
DARPA, the Defense Advanced Research Projects Agency, again hosted its Grand Challenge, an event that pits autonomous automobiles against nature, time and each other. The two previous races, in 2004 and 2005, were held on a desert course, with the goal for each "bot" simply to get from start to finish. This year, DARPA ramped up the level of complexity by presenting bots with an array of "urban challenges." Traffic, four-way stops and other robots combined to make this year's event far more difficult.
Initially 35 teams were entered. Some came from universities, others from defense contractors, and a few were primarily private endeavors. After a week of qualification events, the field was pared down to 11 robots that would compete in the final event on Nov. 3.
"This was really a different event and a very difficult event," Tony Tether, director of DARPA, told the teams prior to announcing the finalists. "And I know, with maybe a very few exceptions, all of you believe that you should be in the finals. Unfortunately that's not going to be the case."
Most of the bots that didn't qualify were guilty of violating California traffic laws, such as making unsafe maneuvers or running stop signs. There were, however, a few minor collisions, generally between bots and barricades. During the qualification trials, speeds tended to remain under 20 mph, so damage was limited when collisions did occur.
In the final event, several never-before-attempted feats were undertaken. First, in several spots, the bots traveled "at speed" -- velocities similar to those of human drivers. The final event also marked the first time bots faced off head to head, be it at an intersection or passing a slower bot on the road. Adding to the challenge was a fleet of human drivers, making the bots' task even more difficult.
All of this took place at the Southern California Logistics Airport -- formerly George Air Force Base. Thanks to its many abandoned housing units and neighborhoods, the site was perfect to test the bots' abilities without risking life or property.
In all, six teams finished the 60-mile final event, far more than anyone had expected. Teams from Stanford, Virginia Tech and Carnegie Mellon took the top three spots, with Carnegie Mellon's bot "Boss" winning the event overall. Stanford's "Junior" took second and a $1 million check, while Virginia Tech's "Odin" came in third and was awarded $500,000.
Despite its remote location, the event drew thousands of spectators and media from around the world. Given that no vehicle even came near the finish line in 2004, many are considering the performances at this year's event a milestone in the development of autonomous machines.
"The 2004 event was equivalent to the Wright brothers' flight at Kitty Hawk, where their airplane didn't fly very far but showed that flight was possible," Tether said prior to announcing the winner. "I believe that the significant progress after 2004 was due to the fact that the community now believed that it could be done."
For more on the DARPA Grand Challenge, look for our special report appearing soon on GTtv.
All over the country, community leaders are looking to boost economic development through various initiatives. One key element in many of those initiatives is the use of information technology. When local governments build IT infrastructure, create e-government applications, assist high-tech startups or otherwise focus on technology, they create conditions that draw businesses to their communities and help retain skilled workers. This paper discusses and provides examples of these various ways local government can use technology to ultimately make a community more attractive to businesses, visitors and residents.