How can we ensure the safety of failure diagnosis systems? This paper proposes a safety assessment model for software used in failure diagnosis systems. The software is represented using production rules, where conditions indicate system failures and conclusions suggest causes or protective actions. The proposed risk criterion considers the reliability of both hardware and the monitored system, deviating from traditional methods that focus solely on internal software consistency. The paper emphasizes that software for failure diagnosis cannot be optimally designed without considering environmental factors. The study challenges conventional verification and validation methods. The proposed assessment model offers a framework for evaluating software contributions to risk in failure diagnosis systems. Future work may focus on implementing and validating the model in real-world applications.
Published in the International Journal of Reliability, Quality and Safety Engineering, this paper fits directly within the journal's scope of exploring safety and reliability in engineering systems. The proposed safety assessment model for software in failure diagnosis systems aligns with the journal's emphasis on ensuring the dependability of complex systems.