Data fusion and maintenance policies for continuous production processes

Continuous production processes involve round-the-clock operation of several, almost identical pieces of equipment that are required to operate concurrently. The failure of a piece of equipment interrupts the flow of production and incurs losses due to waste of raw material. The incidence of inservice failure can be reduced through preventive maintenance. However preventive maintenance also interupts production and creates waste. Thus, the desire to prevent in-service failures while minimizing the frequency of preventive maintenance gives rise to the problem of determining an optimal system-wide maintenance interval. This paper proposes a procedure for addressing problems of this type. The procedure requires as input two quantities: a probability model for the failure of equipment, and a utility function which describes the consequences of scheduled and unscheduled stoppages. The proposed failure model is based on expert opinion and the pooling (or fusion) of the data from various pieces of equipment. The pooling of data is based on the information content of each data set in the sense of Shannon.