Autonomous and Market-Based Fault Tolerant Algorithms for Multi-Robot Cooperation

Abstract

This paper presents a market-based approach for multi-robot cooperation. The approach uses auctioneering as the decision-making mechanism for assigning robotic tasks. A task may be carried out either by one robot or cooperatively by multiple robots, depending on the task requirements and the available robotic resources. The system autonomously determines the appropriate number of robots and selects the most suitable robots in the fleet to carry out the task, while accommodating partial or full failure of robots during the task execution, in an unknown, dynamic and unpredictable environment. The system does this efficiently by choosing the candidate robot that best matches the demands of the task. The feasibility of the developed scheme is demonstrated by implementing the approach on a team of simulated mobile robots that transport multiple objects to a goal location.