What is OEE? Formula, worked example and benchmarks

A machine can be "running" all shift and still lose a third of its capacity. It stops for a few minutes at a time, runs a little slower than it should, and makes some parts that are rejected. None of these losses looks serious on its own. OEE adds them up.

The OEE formula

OEE is the product of three factors, each expressed as a percentage:

OEE = Availability × Performance × Quality

• Availability = run time ÷ planned production time. It captures breakdowns, changeovers, waiting for material and every other stop.
• Performance = (ideal cycle time × total count) ÷ run time. It captures slow cycles and small stops too short to be logged.
• Quality = good count ÷ total count. It captures scrap and rework.

A worked example

A CNC machine is planned for an 8-hour shift with a 30-minute break, so planned production time is 450 minutes. It stops for 45 minutes in total. The ideal cycle time is 1.0 minute per part. It makes 360 parts, of which 351 are good.

A quick check: 351 good parts against 450 possible at the ideal rate is also 78.0%. If the two methods disagree, one of your inputs is wrong.

What is a good OEE?

85% is the widely quoted world-class figure for discrete manufacturing. In practice, plants that measure OEE automatically for the first time usually land between 40% and 60%. That is not bad news. It means the capacity you need may already be on your floor.

At a leading tube and cycle manufacturer, accurate OEE data allowed three planned machine purchases to be postponed by two years.

The six big losses

Every OEE loss falls into one of six categories. Naming them correctly tells you who should fix them.

• Availability: equipment failure, and setup and adjustment.
• Performance: idling and minor stops, and reduced speed.
• Quality: process defects, and reduced yield at start-up.

Why manual OEE is usually wrong

Hand-filled logs miss short stops, round downtime to the nearest 15 minutes and are completed at the end of the shift from memory. At one customer, cycle-time benchmarks turned out to be only about 60% accurate once machine data replaced the log sheet. Automatic capture fixes the inputs: run and stop signals come from the machine, part counts from the machine, and the operator only adds the reason.

How to raise OEE

1. Measure automatically on the bottleneck machines first.
2. Code every stop with a reason at the machine, while it is fresh.
3. Work the Pareto: attack the biggest loss, not the most visible one.
4. Review every shift, not every month. Hourly target-versus-actual makes problems visible while they can still be recovered.
5. Link breakdowns to maintenance so every failure becomes a ticket with a root cause.

Typical plants that follow this approach with Efficio OEE software raise OEE by 15–25%.

FAQ