ABSTRACT
Enhancing systems with advanced diagnostic capabilities for detecting, locating, and compensating faults during operation increases autonomy and reliability. To assure that the diagnosis-enhanced system really has improved reliability, we need – besides other means – to check the correctness of the diagnosis functionality. In this paper, we contribute to this challenge and discuss the application of testing to the case of model-based diagnosis, where we focus on testing the system models used for fault detection and localization. We present a simple use case and provide a step-by-step discussion on introducing testing, its capabilities, and arising issues. We come up with several challenges that we should tackle in future research.
