An empirical bayes strategy for analysing manufacturing data in real time

This paper discusses and develops a real-time strategy to monitor manufacturing process control data. The strategy presented in this paper is designed to intercept, analyse and monitor data as measurements are generated from a high-volume information-intensive manufacturing process. Empirical Bayes theory is used to develop a method to monitor and analyse continuous type measurements generated by a manufacturing process in real time. Several specific techniques will be proposed for detecting manufacturing process problems and examples of applying these techniques will be shown.