Fault Management of Robot Software Components Based on OPRoS

Abstract

Component-based robot development has been a vibrant research topic in robotics due to its reusability and interoperability benefits. However, robot application developers using robot components must invest non-trivial amount of time and effort applying fault tolerance techniques into their robot applications. Despite the need for a common, framework-level fault management, the majority of existing robot software frameworks has failed to provide systematic fault management features. In this paper, we propose a fault management method to detect, diagnose, isolate and recover faults based on the OPRoS software framework. The proposed method provides a collective, framework-level management for commonly encountered robot software faults, thereby reducing the application developers' efforts while enhancing the robot system reliability. To verify the effectiveness of the proposed approach, we have implemented a prototype reconnaissance robot using OPRoS components and injected different types of faults. The results of the experiments have shown that our approach effectively detects, diagnoses, and recovers component faults using the software framework.