The Ecolab Clearing Plant outside Chicago, IL had plenty of sensor data but they were not structured for OPE (overall process effectiveness) or OEE (overall equipment effectiveness) analysis. Join us for this session as we walk through how we used the PI System with function primitives and other Asset Framework and Event Frames capabilities for data modeling, and then applied the manufacturing context to raw sensor (time-series) measurements and lab quality data to prepare for OPE diagnostics. This upfront data engineering also proved to be critical for PI Vision and business intelligence displays for visual analytics. Finally, we will discuss the insights and business benefits we have seen so far, including the next phase in this ongoing work to incorporate machine learning for additional analytics.