I wanted to share our implementation of real-time OEE analytics feeding directly into Opcenter’s Advanced Planning and Scheduling module. We previously calculated OEE manually at shift end, which meant our APS schedules were always based on yesterday’s performance data. This created a disconnect between planned capacity and actual shop floor reality.
Our goal was to stream live equipment performance from SCADA into Opcenter 4.1 and have APS dynamically adjust production schedules based on real-time OEE trends. The implementation took about 6 weeks and involved integrating OPC UA data from 18 production lines with the Advanced Planning module.
The results have been impressive - we’ve improved our overall OEE from 68% to 79% and reduced schedule misses by 34%. The key was mapping live SCADA events to OEE components and creating feedback loops that trigger schedule recalculations when performance deviates from targets.
Event classification was definitely our biggest challenge. We created a hybrid approach - about 70% automated classification and 30% operator validation. Each production line has specific PLC tags that indicate machine state (running, stopped, setup, etc.) and reason codes. We built a mapping table in Opcenter that translates these state combinations into OEE event categories.
For example, if Machine.State=Stopped AND Machine.ReasonCode=0, it’s classified as an unplanned stop (availability loss). If ReasonCode=10-19, it’s planned maintenance (excluded from OEE). The operators have a shop floor terminal where they can correct misclassifications within 15 minutes. After the 15-minute window, the automated classification locks in. This gives us both speed and accuracy.
This sounds exactly like what we need. Can you share more details about the OPC UA data streaming setup? Specifically, how did you handle the data volume from 18 production lines without overwhelming the Opcenter database? We’re concerned about performance impact if we start streaming real-time equipment data.
Great question on data volume management. We implemented an edge aggregation layer that pre-processes SCADA data before sending it to Opcenter. Raw PLC data comes in every 100-500ms, but we aggregate it to 1-minute intervals for availability and 5-minute intervals for quality metrics. This reduces the data stream by about 95% while still maintaining enough granularity for meaningful OEE calculations. The aggregation happens on dedicated edge servers running custom OPC UA clients, so the load never hits the Opcenter server directly.