Description
Since the pioneering work of Walter A. Shewhart in the 1920s, process monitoring has been growing in importance and is currently acknowledged as a key activity in process operations. As process monitoring approaches its 100 years of existence, it is possible to recognize the existence of several evolutionary trends during this extensive period of time that shaped the nature of many solutions and methods proposed. Some of these trends are well-known, while the existence of others is not so well-perceived and appreciated. In this talk, an overview will be provided for several of this old and new trends, as well as examples illustrating their current progress. Among the trends addressed, are:
-
From univariate, to multivariate, to high-dimensional (“mega-variate”) systems
-
From stationary, to dynamic, to non-stationary processes
-
From detection, to diagnosis, to prognosis
-
From monitoring the mean, to dispersion, to correlation