IRIS-01-399

Sold!: Market methods for multi-robot control

Brian Gerkey and Maja J. Mataric

The key to exposing the potential of multi-robot systems is cooperation. How can we achieve cooperation in systems composed of failure-prone autonomous robots operating in noisy, dynamic environments? In this paper, we present a novel method of dynamic task allocation for groups of robots. We have implemented and tested a system which we call MURDOCH, based on a principled, resource-centric, publish/subscribe communication model. MURDOCH produces a distributed approximation to a global optimum of resource usage. We have validated this system in two very different domains: a tightly-coupled multi-robot physical manipulation task and a loosely-coupled many-robot experiment in long-term autonomy.