Networked architecture for multi-robot task reallocation in dynamic environment

Abstract

Distributed and efficiently allocating tasks to networked robots in dynamic environment pose challenge in both theory and application in robotics. A fundamental issue here is how to reallocate tasks during task execution. In this paper, by integrating a high-level market-based algorithm with a low-level mobile ad-hoc network communication structure, we propose a new cooperation architecture named Market-MANET (M-M) architecture, for networked robot task allocation. Additionally, an adventure mechanism is introduced into the M-M architecture to increase the success probability in accomplishing tasks. The effectiveness of the proposed architecture is verified by simulations of task transportation.