Fourth International Conference on Autonomous Agents and Multi Agent Systems Workshop on Learning and Adaptation in Multi Agent Systems

TBA Peter Stone Computer Science Department, University of Texas at Austin, USA Multi-Robot Learning for Continuous Area Sweeping Tasks Abstract: Asmobile robots become increasingly autonomous over extended periods of time, opportunities arise for their use on repetitive tasks. We define and implement behaviors for a class of such tasks that we call continuous area sweeping tasks. A continuous area sweeping task is one in which a group of robots must repeatedly visit all points in a fixed area, possibly with non-uniform frequency, as specified by a task-dependent performance criterion. Examples of problems that need continuous area sweeping are trash removal in a large building and routine surveillance. This talk will begin with a single-robot approach to this problem. We then extend the approach to multi-robot scenarios with a focus on adaptive and decentralized task assignment in continuous area sweeping problems. The proposed method is stable in environments with dynamic factors, such as robot malfunctions or the addition of new robots to the team. The approach is fully implemented and tested both in simulation and on physical robots.