Distributed fault diagnosis for multiple mobile robots using an agent programming language

Abstract

Programming autonomous multi-robot systems can be extremely complex without the use of appropriate software development techniques to abstract away the hardware heterogeneity and to overcome the complexity of distributed software to coordinate autonomous behavior. Moreover, real-world environments are dynamic, which can generate unpredictable events that can lead the robots to failure. This paper presents a highly abstract cooperative fault diagnosis method for a team of mobile robots described through a high level programming environment based on ROS (Robot Operating System) and the Jason multi-agent framework. When a robot detects a failure, it can perform two types of diagnosis methods: a local method executed on the faulty robot itself and a cooperative method where another robot helps the faulty robot to determine the source of failure. A case study demonstrates the effectiveness of out approach on two robots.