Anticipating performance of work stations in MMPs at sensor breakdowns

Multi-Station Manufacturing Processes (MMPs) occasionally encounters the problem of deviation in the attributes of the products as compared to the design specifications. Sensors are installed in the work stations to detect the sources of errors in the product dimensions. This paper identifies the problem concerned with breakdown of the sensors and proposes an approach that identifies the interdependence relations among the various sensors using Bayesian Networks. Particle Swarm Optimization technique has been used to search the Optimal Bayesian Network. This proposed strategy will aid the manufacturers to check the delay in production time and to control the quality of production at times of sensor breakdown.