ANN ARBOR—Robots developed at the University of Michigan could one day be deployed to warzones to search out and neutralize roadside bombs, thanks to a team of more than 20 engineering students led by Edwin Olson, an assistant professor of computer science and engineering.

Olson and his team were selected as one of six finalists in the 2010 Multi Autonomous Ground-Robotic International Challenge (MAGIC). The competition, open to corporate and academia-based teams, required competitors to develop multivehicle robotic teams that would be able to execute intelligence, surveillance and reconnaissance missions in an unknown urban environment with little or no human supervision.

MAGIC is jointly sponsored by the Australian and U.S. departments of defense, and was created to attract innovative proposals from worldwide research organizations in order to develop next-generation fully autonomous ground vehicle systems that can be deployed effectively in military operations and civilian emergency situations. The final stage of competition will take place in November 2010 at an undisclosed location in South Australia. The winning team will take home a top prize of $750,000, with second and third place teams receiving $250,000 and $100,000, respectively.

The U-M team was shortlisted in November 2009 as one of 12 semifinalists in the competition, and was one of 10 teams to be awarded $50,000 to aid its research. On Monday, the team was granted an additional $50,000, as were all six finalists.

Olson and his team were elated to hear the announcement.

“I’m very proud of this team’s accomplishments,” he said. “We’ve done a great job at building a system that can process sensor data and make decisions autonomously, which in turn allows us to field more robots at the same time without the operators going crazy. And more robots will give us an advantage when it comes to completing the missions quickly at the finals in November.”

Olson’s team conducted previous tests with five robots and plan to field up to 15 in the final round.

The six MAGIC finalists were selected after representatives from the U.S. Department of Defense and the Australian Department of Defence Science & Technology Organisation made site visits to assess each team. The evaluators observed the robots as they performed a variety of tasks including accurately mapping the indoor/outdoor challenge area, as well as correctly locating, classifying and recognizing all simulated threats, including differentiating between “good guys” and “bad guys” based on the color of their clothing. The real-world application of this would be to make military robots more autonomous.

“On military field situations today, five or six people are often required to control one robot,” Olson said. “The goal here is to switch that ratio, so you have two or three humans overseeing a team of 20 or 30 robots. We want to push the autonomy onto the robots and have them do more cooperation and collaboration, rather than having the humans do all the thinking.

“One of the things that really excites me about autonomous robotics is that, by removing people from harm’s way, there is less pressure to use lethal force. In the end, I believe that robots like these will not only save the lives of our soldiers, but will help save the lives of civilians caught in dangerous places.”

The U-M team’s approach to the MAGIC challenge has been to develop and deploy a large field of low-cost robots from mostly commodity components, and to leverage cognitive science research at Michigan to imbue this robotic team with as much autonomy and collaborative intelligence as possible. The result is a team of reliable robots that works together to execute missions, even as some robots are “taken out” by judges during the competition.

If eventually deployed into the field of battle, the robots could be armed with technology that to disarm improvised explosive devices, something that was only simulated during the MAGIC competition. These types of vehicles could also be used to conduct search and rescue operations in civilian emergencies.

The U-M team is leveraging support from Soar Technology, the cognitive science company founded from research conducted by John Laird, a professor in the Department of Electrical Engineering and Computer Science.

The University of Michigan College of Engineering is ranked among the top engineering schools in the country. At $160 million annually, its engineering research budget is one of the largest of any public university. Michigan Engineering is home to 11 academic departments and a National Science Foundation Engineering Research Center. The college plays a leading role in the Michigan Memorial Phoenix Energy Institute and hosts the world class Lurie Nanofabrication Facility. Find out more at www.engin.umich.edu.

MAGIC ChallengeU-M MAGIC teamEdwin Olson’s APRIL LaboratoryVideo of robots in action