Quality risk analysis at sampling stations crossed by one monitored product and an unmonitored flow

When inspection economies are implemented in multi-product, multi-stage, parallel processing manufacturing systems, there exists a significant risk of losing control of the monitoring efficacy of the sampling strategy adopted. For a product-based sampling decision limited to a particular station in a production segment, the randomness of the departure process and the merging of different product flows at the machines of the different stations subvert the regularity of deterministic sampling. The risk of not regularly monitoring any machine in the segment can be measured in terms of maximum number of consecutive unsampled items. In this study, the distribution of this measure at sampling station machines is developed for a production scenario characterized by one monitored product and an unmonitored flow and compared with the behavior of the same measure at non-sampling station machines. The prediction models illustrated prove fundamental pragmatic tools for quality management involved in sampling strategy-related decisions.