Performance Modelling and Analysis of Unreliable Manufacturing Cells with Multiple Rework Attempts and Variable Defect Probabilities

This paper addresses the problem of accurate assessing the performance measures of manufacturing cells prone to random breakdowns, multi-rework, and scrap. To analyse the dynamic and stochastic behaviour of the manufacturing cell, we propose a new, realistic, and general analytical model based on Markov chain approach. The proposed approach applies for any specific and allowed number of rework attempts as pre-specified by production and quality departments. The probabilities of product acceptance, rejection, and scrapping depend on the number of reworks the part has undergone. Several analytical formulations are developed, according to several industrial contexts, to assess the manufacturing cell KPIs such as throughput, effectiveness, availability, quality, and utilization rates, number of raw, good, and scrapped parts, and the manufacturing cell yield. Simulation models were also developed, and hundreds of cell configurations were explored to validate the proposed analytical models. A comparative analysis shows that the recommended strategy outperformed other strategies that were considered in the literature and in the real world.