Time series analysis of process data

Statistical Process Control (SPC) is an integral component of almost every industrial process, and proper outlier (i.e., out of control) detection is crucial if processes are to remain in statistical control. The goal of this research is to determine whether a simple model may be useful as an approximation to a more exact and thus more difficult model; and still provide sufficient accuracy in outlier detection. We test an ARMA (1,1) model with the Chen and Liu Joint Estimation (JE) outlier detection algorithm with different sets of process data. We find that this approach is quite useful, especially for practitioners.