On multistage statistical process control

Most products and services today are the results of several process stages and steps. With the current emphasis in industry on improved quality, control charts are widely used in process monitoring. However, conventional statistical process control (SPC) techniques focus mostly on individual stages in a process and do not consider disseminating information throughout the multiple stages of the process. Such techniques are ineffective in analyzing multistage processes. The cause-selecting chart (CSC), based on output adjusted for the effect of the incoming quality is an effective SPC tool for analyzing multistage processes. This paper discusses several important problems associated with conventional CSCs. First, the model relating input and output measures often needs to be estimated in practice before the CSC is implemented. Little is known about the performance of the CSC when the model parameters are estimated. Second, the simple linear regression model widely discussed in the CSC is insufficient to capture the stochastic behavior of the output. In practice, the process disturbance in a multistage process can be autocorrelated. To deal with autocorrelated disturbance, the cumulative score (Cuscore) chart and the triggered Cuscore chart are proposed. Finally, extensions of the CSC to multivariate CSCs are discussed.