OEE calculation looks simple until you try to do it on a real production line — and find that the definition of "planned production time" is ambiguous, the PLC doesn't expose good counts directly, and your downtime log has gaps. This guide walks through each calculation step with real numbers, and flags the mistakes most plants make.
OEE (Overall Equipment Effectiveness) is the product of three components:
Each component is a percentage (0–100%). Multiply them together. A machine with 90% Availability, 95% Performance, and 98% Quality has OEE = 0.90 × 0.95 × 0.98 = 83.8%.
Availability = fraction of scheduled production time that the machine was actually running (not stopped).
Planned Production Time = total shift time minus any scheduled stops (lunch breaks, planned maintenance windows, planned changeovers — whether to include changeovers in Planned Production Time vs. treating them as downtime depends on your convention).
Downtime = sum of all unplanned stops during the shift (machine faults, material shortages, operator absences, etc.)
Shift: 8 hours (480 minutes)
Scheduled break: 30 minutes → Planned Production Time = 450 minutes
Unplanned downtime events: 22-minute conveyor jam + 13-minute sensor fault = 35 minutes
Availability = (450 − 35) ÷ 450 = 415 ÷ 450 = 92.2%
Common mistake: including changeover time in "downtime" without also including it in Planned Production Time. If changeovers are planned, they should be excluded from both — or included in both. Inconsistency here creates OEE above 100% or below realistic values.
Performance = how efficiently the machine ran during its available run time, expressed as actual throughput vs. theoretical at ideal speed.
Ideal Cycle Time = the fastest sustainable cycle time the machine can achieve at its nameplate design speed. This should come from the machine specification, not from "best shift" data.
Total Count = all parts produced (both good and reject) during Run Time.
Run Time = Planned Production Time minus Downtime (same as the numerator of Availability calculation above).
Run Time = 415 minutes
Ideal cycle time = 0.6 minutes/part
Theoretical max output = 415 ÷ 0.6 = 691.7 parts
Actual output (total count including rejects): 642 parts
Performance = (0.6 × 642) ÷ 415 = 385.2 ÷ 415 = 92.8%
Common mistake: using "average actual cycle time" as the Ideal Cycle Time. This artificially inflates Performance toward 100% because you're measuring against your own average rather than the machine's design capability. Always use the specification-rated cycle time.
Quality = fraction of total parts produced that passed first inspection.
Good Count = parts that passed first-pass inspection without rework
Total Count = all parts produced including scrap and rework
Total Count = 642 parts
Scrap: 8 parts; Rework: 5 parts (counted as non-conforming)
Good Count: 642 − 8 − 5 = 629 parts
Quality = 629 ÷ 642 = 97.9%
Availability = 92.2%
Performance = 92.8%
Quality = 97.9%
OEE = 0.922 × 0.928 × 0.979 = 83.8%
This is a good OEE for a discrete manufacturing line. World-class is 85%+. The primary improvement opportunity here is Performance (92.8%) — investigate what is causing the machine to run slower than its ideal cycle time.
Manual OEE calculation is a starting point, not a production system. The problems: operator data entry errors, missed micro-stops, rounded downtime, delayed reporting (next day at best). Automated OEE from OPC UA machine data is the standard in lean manufacturing — every cycle time, fault event, and production count captured automatically, OEE calculated per hour in real time.
Automate this calculation → Shopfloor Copilot calculates OEE from your OPC UA machine data in real time. No spreadsheets, no manual entry.
See OEE Monitoring →