What is OEE?
OEE (Overall Equipment Effectiveness) measures how much of the time you planned to make product was actually spent making good product at full speed. A line at 100% would run every planned minute, at its ideal rate, with no rejects. Nobody gets there, and that's fine. OEE is useful because it splits the shortfall into three losses you can do something about:
- Availability: time lost to stops, including breakdowns, waiting and changeovers.
- Performance: time lost to running slower than ideal, including short stops and minor jams.
- Quality: time spent making product that ends up as scrap, rework or rejects.
How to calculate OEE
- Planned production time = shift length − planned breaks
- Run time = planned production time − unplanned downtime
- Availability = run time ÷ planned production time
- Performance = (ideal cycle time × total count) ÷ run time
- Quality = good count ÷ total count
- OEE = availability × performance × quality
Worked example Example data
A fictional packing line, used for teaching only. Not a real site or client.
| 8-hour shift (480 min) minus 30 min of breaks | 450 min planned |
| 45 min of unplanned downtime, including a 25-min changeover | 405 min run time |
| Availability: 405 ÷ 450 | 90.0% |
| Ideal rate 60 units/min (1 second each). 20,000 made. Performance: 20,000 ÷ (405 × 60) | 82.3% |
| 19,400 good. Quality: 19,400 ÷ 20,000 | 97.0% |
| OEE: 90.0% × 82.3% × 97.0% | 71.9% |
A useful cross-check: 19,400 good units at 1 second each is about 323 minutes of fully productive time. 323 ÷ 450 planned minutes is the same 71.9%. On this example line, slow running costs more time (about 72 minutes) than all the stops put together (45 minutes), which you'd never spot from the downtime sheet alone.
Common OEE mistakes
- Treating changeovers as planned time. If you take changeovers out of planned production time, they vanish from availability and the line looks better than it is. Changeovers are a loss you can shrink (that's what SMED is for), so count them as downtime.
- Using the "usual" speed as the ideal rate. If the ideal is set to what the line normally does, performance sits near 100% and the speed loss is hidden. Use the designed or validated top speed of the bottleneck. A performance figure above 100% means the ideal is set too low.
- Counting rework as good. Product that had to be re-packed, re-labelled or re-checked wasn't right first time. Leave it out of the good count.
- Averaging OEE across lines. A simple average of percentages hides the big lines. Add up the minutes and counts across lines first, then calculate.
- Chasing the number instead of the losses. OEE is a pointer. The value is in knowing which loss is biggest and why.
What does a good OEE look like?
The figure you'll see quoted most is 85% as "world class", usually traced back to Seiichi Nakajima's work on Total Productive Maintenance. It's often broken down as roughly 90% availability, 95% performance and 99.9% quality. Treat it as a commonly cited benchmark, not a rule. It came from discrete, high-volume manufacturing, and lines with frequent changeovers, cleans and short runs are a different game.
The comparison that matters is your line against itself, measured the same way each week. If OEE goes up because a changeover got shorter or a jam was designed out, that's real. If it goes up because someone moved the definitions, it isn't.
