ON THE INTRODUCTION OF A RESTART POLICY IN ANALYTICAL MODELS OF MANUFACTURING SERIAL LINES

This paper deals with analytical models of manufacturing serial lines where adjacent line stages (i.e. machines or work-stations) are decoupled by a finite buffer. Specifically, the aim is to take into account the inherent uncertainty arising from unpredictable, or predictable but not constant, disruptions of the manufacturing process. As regards production lines, since times to failure and times to repair can generally be assumed to be exponentially (or geometrically) distributed, Markov process models are suitable for performance assessment. On the other hand, Bernoulli process models are more appropriate in case of assembly lines. This work discusses some of the most popular Markov and Bernoulli models of manufacturing lines; then, an interesting extension, called “restart policy”, is presented. The new models with restart policy can be applied to a number of industrial production and assembly systems when it is important to reduce the probability of blocking events.