Robotics has progressed quick in recent decades, but the focus has been on single robots performing single tasks.
The hurdle has been attempting to get groups of robots to work as a team to perform tasks together. Long the bane of robotics, researchers at MIT are trying to get groups of robots, perhaps numbering in the thousands or more, to perform large and complicated tasks that a single robot can't do by itself.
By focusing on the software and networking components of the equation, they hope to solve the problem.
"In [multi-agent] systems, in general, in the real world, it's very hard for them to communicate effectively. If you have a camera, it's impossible for the camera to be constantly streaming all of its information to all the other cameras. Similarly, robots are on networks that are imperfect, so it takes some amount of time to get messages to other robots, and maybe they can't communicate in certain situations around obstacles," said Christopher Amato, a post-doctoral student at MIT's Computer Science and Artificial Intelligence Laboratory, in a statement.
Software that mimics human-like decision making it difficult to write for even a single robot performing a specific task, and as the complexity of the system increases - either by adding in robots or making the task or environment more varied - the hurdles become even higher.
"When you try to make a decision, there's some uncertainty about how that's going to unfold. Maybe you try to move in a certain direction, and there's wind or wheel slippage, or there's uncertainty across networks due to packet loss. So in these real-world domains with all this communication noise and uncertainty about what's happening, it's hard to make decisions," said Amato.
By breaking up the functions into three different inputs, MIT researchers hope to solve this problem. A set of algorithms controls the robots, individually or as a group; the second input focuses on the execution of programs in specific environments; and the third controls the valuation of the given tasks.
The findings of MIT's research will be presented in May 2014 at the International Conference on Autonomous Agents and Multiagent Systems in Paris.
