Improving quality in flexible manufacturing systems: A bottleneck transition approach

In this paper, a Markov chain model is introduced to study product quality in flexible manufacturing systems with batch productions. Using this model, we introduce good and defective states as the cycles where good quality and defective parts are produced, respectively. Then the product quality is a function of the transition probabilities characterizing the changes among good and defective states. To improve product quality, identifying the transition that has the largest impact on quality is important, and such transition is referred to as the quality bottleneck transition (BN-t). Analytical expressions to quantify the sensitivity of quality with respect to transition probabilities are derived, and indicators to identify the BN-t based on the data collected on the factory floor are developed. Through extensive numerical experiments, it is shown that such indicators have high accuracy in identifying the correct bottlenecks and can be used as an effective tool for quality improvement effort. Finally, a case study at an automotive paint shop is presented to illustrate the applicability of the method.