Diagnosing intermittent faults
Almost all approaches to model-based diagnosis presume the system being diagnosed behaves non-intermittently, and its behavior analyzed over a small set (often one) instant. In this paper we show how existing approaches to model-based diagnosis can be extended to diagnose intermittent failures, as they manifest over large numbers of time instants. In addition, we show how to insert probe points which best distinguish among intermittent faults which explain the symptoms.
de Kleer, J. Diagnosing intermittent faults. Twenty-Second AAAI Conference on Artifical Intelligence (AAAI-2007); 2007 July 22-26; Vancouver; BC; Canada.