Algorithmic Statistical Process Control: An Elaboration

Abstract

Statistical process control (SPC) has traditionally been applied to processes in which successive observations would ideally be independent and identically distributed as a basis for achieving fundamental process improvement. Stochastic control, on the other hand, addresses situations in which observations are dynamically related over time; its intent is to run the existing process well, as opposed to improving it as such. A schema is presented for uniting traditional SPC and feedforward/feedback control into a system that exploits the strengths of both, Building on past work by MacGregor, Box, Åström, and others, we discuss the theory and practice of such an approach, along with a consideration of research and technical issues that arise.