Students at MIT’s Computer Science and Artificial Intelligence Laboratory are using new algorithms and lateral thinking to enable household robots to work around their physical shortcomings. Unlike an industrial setting, where positioning make objects easier to grasp, household robots would need to be able to manipulate objects of any shape in any location.
The MIT research is contained in two, linked, papers. The first concentrates on picking, the other on placing.
A picking algorithm enables a robot to push the object across a table so that part of it hangs off the edge, where it can be grasped. Instead of the commonly used rapidly exploring random tree motion algorithm, which maps out collision-free trajectories through the robot’s environment, the researchers sought a concise way to represent the physical properties of the object to be pushed — how it would respond to different forces applied from different directions. This resulted in a more efficient method of searching.
The other paper focuses on placing and shows how a two-armed robot can use one of its graspers to steady an object set in place by the other. It seeks to invert the motion-planning task. Rather than identifying paths that avoid collisions, it identifies those that introduce them and seals them off.
The research will be presented at the this year’s IEEE International Conference on Robotics and Automation, in Karlruhe, Germany, 6-10 May.
The MIT News site has more info.