A new fault tolerance method for field robotics through a self-adaptation architecture

Abstract

Fault tolerance is increasingly important for urban search and rescue robotic systems because any failure mode may affect the reliability of the robot in meeting mission objectives. To support convenient development of fault tolerant robotic systems, this paper proposes ReFrESH, which is a self-adaptive framework that provides systemic self-diagnosis and self-maintenance mechanisms in the presence of unanticipated situations. Specifically, ReFrESH augments the port-based object by attaching performance evaluation and estimation methods to each functional component so that the robot can easily detect and locate faults. In conjunction, a task level decision mechanism interacts with these fault detection elements in order to dynamically find an optimal solution to mitigate faults. A demonstrative application of ReFrESH illustrates its applicability through a task for visual servoing to a target deployed on a multi-robot system.