On improving failure mode and effects analysis ( FMEA ) from different artificial intelligence approaches

This study describes the Failure Mode and Effects Analysis (FMEA) from the point of view of different artificial intelligence approaches. After discussing the main drawbacks of the traditional methodology and summarize the main techniques recommended for its improvement in the recent literature, three techniques of Artificial Intelligence are compared: a fuzzy inference system (FIS), a case reasoning based method (CBR) and a vector support machine based method (VSM). From the results of this study we conclude that the best approach to properly classify the causes of risk of a system or service is the fuzzy inference system, method that, in addition, allows to overcome most of the drawbacks associated with the traditional methodology.