MODEL-BASED REASONING FOR FAULT-TOLERANT ROBOT HARDWARE AND SOFTWARE
Michael Hofbaur, Gerald Steinbauer, Franz Wotawa · 2015
Abstract: An autonomous mobile robot comprises a number of complex software and hardware modules and interactions between them. Because of this complexity there is always the possibility of faults at runtime even in the case of well-defined and rigorously executed design processes. If an autonomous robot is unable to automatically detect and repair such faults it tends to fail to finish its task. In this paper we present model-based reasoning techniques which enable a robot to perform such supervision tasks. We explain in detail the methods for software and hardware supervision. Finally, we present examples for the application of the methods to a real robot system.