Comments on linear circuit fault diagnosis using neuromorphic analyzers

In a recent paper, Spina and Upadhyaya presented a method for the fault diagnosis of analog linear circuits. The method, which is based on a white noise generator and an artificial neural network for response analysis, has been applied to circuits of reasonable dimensions, taking into account the effect of the component tolerances. However, the proposed method does not take into account the testability analysis of the circuit under test. Research on testability analysis of linear circuits has been developed by several authors in the last 20 years, and algorithms and programs for testability evaluation have been presented in several publications. It is our opinion that the testability analysis concept could be useful in the approach proposed by Spina and Upadhyaya to improve the quality of the results even further. In this brief, we discuss this possibility.