USING ONLINE PROCESS DATA TO IMPROVE QUALITY - CHALLENGES FOR STATISTICIANS

In the process industries measurements on a large number of process va riables are routinely collected at regular intervals by on-line comput ers. This paper makes a case for incorporating these process variables into Statistical Process Control (SPC) schemes. Multivariate statisti cal methods such as Principal Component Analysis (PCA) and Partial Lea st Squares (PLS) can be used to project these data down into low dimen sional spaces where analysis, monitoring and diagnosis are easily perf ormed. Strong justifications for taking this approach are presented an d examples are given, The statistical process control community has be en slow in adapting to the data explosion brought about by the compute r era, It has continued to stick with traditional control charts on th e quality variables and ignored this rich source of additional informa tion on the process. This paper explores some of the reasons for this and argues that the SPC community must adapt rapidly or lose control o f the field to scientists and engineers. The paper also tries to induc e statisticians into looking more seriously at the many unsolved probl ems in this area of reduced rank multivariate statistics.