Simulation of Fault Detection for Robot Applications
As robots are becoming increasingly useful in various application areas, more robot application systems are structured with robot components. Robot components may be used to develop intelligent robot applications that are usually safety or mission-critical. In spite of the many advantages of robot components, component-based robot applications still need to be reliable against runtime faults or failures, such as unanticipated events or component defects. However, current approaches or platforms for supporting component-based robot applications do not address the fault or failure management. This research is to develop a simulator for detecting faults in the unmanned ground vehicle (UGV) application that is composed of multiple robot components. A UGV application is implemented using Microsoft Robotics Developer Studio, and the constituting components for the UGV contain fault detection functions such as sensor data range check or data period failure check. The simulator checks if each component can detect faults encountered while the UGV drives on a simulated road environment.