Three parts, one score
Availability is running time versus planned time. Performance is actual output versus ideal speed. Quality is good units versus total units. OEE multiplies them. A world-class line might sit near 85% — that benchmark is less useful than knowing which of the three is dragging your shift.
Reading the score on the floor
When cost per unit, overtime or CapEx rises, OEE is a sorting tool. Chronic micro-stops and changeovers show up in availability. A line that “runs” but never hits ideal cycle shows up in performance. Escape defects show up in quality. Fix the bucket that moves the bill — not the dashboard colour.
Keeping the number honest
Pick one asset and one shift definition. Capture planned time, downtime, ideal cycle, total count and rejects. Then: Availability = run ÷ planned; Performance = (ideal cycle × total count) ÷ run; Quality = good ÷ total; OEE = the product. PLC, sensor or MES counts beat clipboards. Publish the loss categories next to the percentage, or the percentage becomes decoration.
Key takeaways
- OEE = Availability × Performance × Quality
- The loss Pareto matters more than the headline %
- Same definition, same line, over time
- Automate counts where you can — sheets drift
Frequently asked questions
Is 85% OEE the target for every plant?
It is a common discrete-manufacturing reference, not a law. Improve against your baseline and the losses you can see first.
How often should OEE be calculated?
Shift or daily for the line team; weekly or monthly for management. Sub-shift windows need automated data.
Does OEE apply outside factories?
Classical OEE assumes a defined ideal cycle and good/reject counts. Logistics and facilities use related ideas, usually under different names.