An algorithm for the estimation of minimal cut and path sets from field failure data

Abstract

The problem of analyzing the field failure data generated through non-repairable systems is age old and faced by all industries. Current practice, in most of the industries, is to disassemble the failed system, and observe the failed components. Based on such a list generally percentages are computed to summarize the data and technical actions are taken on a component which has higher percentage of failures, without paying any attention to the technical configuration of the system. Such an approach may be misleading. It is known that a system as a coherent structure of the components, fails as soon as all the components in a specific cut set fail. In this paper, we propose an algorithmic approach to identify minimal cut sets of the system which in turn identifies the critical cut set where technical action is necessary. We establish the consistency of the estimators of minimal cut sets and path sets. Lastly, the algorithm is illustrated by a suitable example.