Monitoring and diagnosing a two-stage production process with attribute characteristics

Multistage process monitoring has recently attracted notable attention in that the statistical relationships between quality variables are taken into account. Here, we dealt with the problem of monitoring and diagnosing a two-stage production process with attribute characteristics in which the outgoing quality variable is impacted by the incoming quality variable from the first process stage. Based on a sampling procedure which inspects each n produced items, these attribute characteristics are assumed to follow binomial and Poisson distributions. Several monitoring techniques including a new method based on the generalized Poisson distribution are presented and the comparison is made to evaluate the effectiveness of these procedures. Moreover, some fault diagnosis methods are fully explored in order to alleviate the identification of the process stage responsible for the outof- control conditions. The results of the simulation based studies reveal that a combined approach consisting of a proportion defective control chart and an adjusted control chart is quite efficacious in addressing the problems with regard to both the detection power and the fault diagnosis.