Autonomous robots could collaborate on mapping hazardous environments
By Ben Coxworth
May 17, 2011
Imagine you're a firefighter arriving at a burning building, but you have no idea what the interior layout of that building is. Do you just enter, then risk your life by randomly walking up and down smokey hallways looking around? What would work better would be if someone or something could quickly map out the building first, then you could take that information and concentrate on getting directly where you need to go. That's the idea behind the U.S. Army's Micro Autonomous Systems and Technology (MAST) Collaborative Technology Alliance Program, which would see swarms of small rolling, hopping, crawling or flying robots working together on reconnaissance missions in civilian or military applications.
The five-year MAST program is led by BAE Systems, and includes a number of American universities.
In a recent experiment, a group of one square foot (929 sq. cm.)-sized wheeled autonomous robots were used to explore and map out an office building. The robots featured navigation technology developed at the Georgia Institute of Technology, with network technology from the University of Pennsylvania, and vision-based techniques from the California Institute of Technology/Jet Propulsion Laboratory (JPL).
A video camera on each robot identified doorways and windows, while a laser scanner measured walls. An onboard inertial measurement unit let them keep track of their movement, and aided in stability. A graph-based system called SLAM (simultaneous localization and mapping) allowed each robot to build a local area map using its sensory data, while also keeping track of its own location within that map.
The robots worked entirely without human input, communicating with one another to coordinate their movements – if two robots were to arrive at an intersection, for instance, they could decide which one would go left and which would go right. Although there was no "lead robot," individual units could recruit others to explore areas as needed.
At the end of the exercise, a detailed floor map was transmitted to the nearby human research team.
In the future, however, it is hoped that scouting robots could also be equipped with tiny wall-penetrating radar units, infrared sensors that would detect anything giving off heat, and flexible "whiskers" that would allow them to sense walls in complete darkness. It is also hoped that the robots will end up being small enough to fit on a person's open palm.
"When first responders – whether it's a firefighter in downtown Atlanta or a soldier overseas – confront an unfamiliar structure, it's very stressful and potentially dangerous because they have limited knowledge of what they're dealing with," said Georgia Tech's Prof. Henrik Christensen. "If those first responders could send in robots that would quickly search the structure and send back a map, they'd have a much better sense of what to expect and they'd feel more confident."
Other members of MAST include North Carolina A&T; State University, the University of California Berkeley, the University of Maryland, the University of Michigan, the University of New Mexico, Harvard University, the Massachusetts Institute of Technology, and Daedalus Flight Systems.