Bayesian networks for nonconformities management

This paper explains BNs’ potentialities as a support tool for graphic and intuitive representations of any process and in particular as an intelligent system: starting from a set of data, BNs are not just an inferential tool but their added value consists of the possibility to encode human knowledge and experience in the model. Thus, once created, a BN can be continuously improved and refined, making the model closer and closer to reality. Bayesian Network are often used for reliability and maintenance systems, but in fact they can be applied to any field. In the present case study BN have been applied to nonconformities management, resulting in an expert system that that can lead improvements both from an operative and a strategic side.