A study of self-diagnosis system for an autonomous mobile robot: coping with self-condition using internal sensory information

When the robots work autonomously in a real environment, it is a required function to recognize self-condition for reliable task execution. Therefore, we discuss and develop a self-diagnosis system and adaptive behavior generation for autonomous robots. We propose a method that copes with self-condition using internal sensory information by many sensors in the system. We focus on functional units in a single robot system and also examine the method which makes the faulty condition divide into three levels. The behavior, which copes with the faulty condition, is set to each level to continue to execute the task. We show experimental results using an omnidirectional mobile robot with self-diagnosis system and our proposed method.