Multivariate Statistical Process Control Applied to an Industrial Production Facility

Abstract This paper addresses the development of a Multivariate Statistical Process Control (MSPC) based performance monitoring system for a high speed polyester film production facility. The plant is a modern film production facility where over six hundred process measurements are recorded at one minute intervals. The objective for applying MSPC was to improve process consistency, detect process disturbances and increase operator awareness of events. The background to MSPC is briefly described and the various stages in the development of the at-line implementation of MSPC for the production line are outlined. Three case studies are used to illustrate the power of the methodology in contrasting situations