The following is from the notes for the hands-on lab "Fit for Purpose - Layers of Analytics using the PI System – AF, MATLAB, Machine Learning..." at PI World 2018, San Francisco, CA. Lab manual and slides are attached - scroll to the end of the blog post.
Briefly, the talk and the lab walks through - with examples - the Layers of Analytics approach which depends on the use case - operations, maintenance, quality etc.
In plant floor process operations, the often used "Descriptive, Diagnostic, Predictive, Prescriptive" layer is explored using an oil refinery unit operation (Alkylation Feed Dryer) example.
In asset maintenance/reliability, the UbM, CbM, PdM layer is explored using a motor/pump assembly with degrading health (inferred via vibration measurements) and predicting RUL (remaining useful life).
For an extended blog post, see Fit for Purpose - Layers of Analytics using the PI System