How to calculate OEE, with a worked example
OEE is Availability × Performance × Quality. Here is how to calculate each one from a real shift, and the mistakes that make OEE numbers meaningless.
· 2 min read
Overall equipment effectiveness (OEE) tells you how much of your planned production time was genuinely productive: making good parts, as fast as the process allows, with no stops. It is the most widely used single measure of manufacturing productivity, and it is simple to calculate once the inputs are agreed.
The formula
OEE is the product of three factors:
OEE = Availability × Performance × Quality
- Availability is run time divided by planned production time. It captures stops: breakdowns, changeovers, waiting for material.
- Performance is how fast the machine ran while it was running, compared with its ideal cycle time. It captures slow running and short stops.
- Quality is good parts divided by total parts. It captures scrap and rework.
A worked example
Take one eight-hour shift on a single machine.
| Input | Value |
|---|---|
| Shift length | 480 minutes |
| Planned breaks | 30 minutes |
| Unplanned downtime | 45 minutes |
| Ideal cycle time | 30 seconds per part |
| Total parts made | 700 |
| Rejected parts | 28 |
Planned production time is the shift minus planned breaks: 480 − 30 = 450 minutes.
Run time is planned production time minus downtime: 450 − 45 = 405 minutes.
Availability = 405 ÷ 450 = 90.0%
Performance = (ideal cycle time × total parts) ÷ run time = (0.5 min × 700) ÷ 405 = 350 ÷ 405 = 86.4%
Quality = good parts ÷ total parts = 672 ÷ 700 = 96.0%
OEE = 0.900 × 0.864 × 0.960 = 74.7%
You can check the result another way: OEE is also good parts × ideal cycle time ÷ planned production time, which is 672 × 0.5 ÷ 450 = 74.7%. If your two calculations disagree, one of the inputs is wrong.
Try your own numbers in the OEE calculator.
What is a good OEE score?
85% is often quoted as world class for discrete manufacturing, and many plants that measure honestly for the first time find they are somewhere between 40% and 60%. The absolute number matters less than the trend and the losses behind it. A line at 55% that knows exactly where the other 45% goes is in a better place than one at 80% built on guesses.
Common mistakes
- Using a generous ideal cycle time. If the ideal is set to what the machine "normally" does, performance losses disappear. Use the fastest sustainable rate, from the machine specification or a demonstrated best.
- Leaving changeovers out of planned time. Changeovers are an availability loss. Excluding them hides one of the biggest improvement opportunities.
- Averaging OEE across machines. A simple average of percentages is misleading. Sum the underlying times and counts, then calculate.
- Recording stops from memory at the end of the shift. Short stops are forgotten and reasons become "other". Capture them as they happen.
- Chasing the number instead of the losses. OEE is a pointer. The value is in the stop reasons, scrap causes and slow-running patterns behind it.
Getting reliable data
The calculation is the easy part. The hard part is collecting accurate run time, counts and stop reasons without adding work for operators. That usually means taking counts directly from the machine and giving operators a quick way to tag stoppages at the line. It is exactly the kind of system we build: see OEE and production systems.