Proceedings Abstracts of the Twenty-Third International Joint Conference on Artificial Intelligence

Capabilities in Heterogeneous Multi Robot Systems / 3207
Jennifer Buehler

The increasing variety of robotic systems create the need for flexible architectures enabling easy integration of new robot configurations into existing multi-robot systems. This requires methods for general reasoning about what different robots are capable of doing. Teamwork is a very important factor in complex, dynamic domains. In heterogeneous teams, robustness and flexibility are increased by the diversity of the robots, each contributing different capabilities. Consequently it is reasonable to explicitly take the robots' capabilities into account when determining which robot is best suited for a task. This work develops a framework that formalizes robots' capabilities, relating to hard- and software configurations and providing a means to estimate a robot's suitability for a task. A learning algorithm for robot capabilities is included.