How to calculate OEE

To calculate OEE, work out three factors, Availability, Performance and Quality, then multiply them together. The maths is simple; the hard part is collecting honest data. This guide walks the full method with a worked shift example, then flags the mistakes that quietly inflate the number.

If you're new to the concept, start with what is OEE. This guide assumes you know what the three factors mean and focuses on computing them.

The data you need

Five inputs give you OEE for any window, whether a shift, a day or a batch:

• Planned production time: scheduled time the line was meant to run (exclude unscheduled time and planned breaks).
• Stop time: total time the line was down during planned time (breakdowns, changeovers, starvation).
• Ideal cycle time: the fastest sustainable time to make one unit (the nameplate rate).
• Total count: all units produced in the window, good and bad.
• Good count: units that met spec the first time, no rework.

Step by step, with a worked example

One 8-hour shift on a single line:

Step 1 - Availability

Run time = planned time − stop time = 480 − 80 = 400 minutes.

Step 2 - Performance

At the ideal cycle time the line could make 24,000 units in 400 minutes of run time. It actually made 19,200.

Equivalently, Performance = (Ideal Cycle Time × Total Count) ÷ Run Time. Performance above 100% means your ideal cycle time is set too slow: fix the standard, don't cap the number.

Step 3 - Quality

Of the 19,200 units, 18,816 were good first time (384 rejects or reworks).

Step 4 - Multiply

Five mistakes that inflate OEE

• Padding planned time. Quietly shrinking scheduled time hides availability loss. Pick one honest definition and stick to it.
• Ignoring small stops and speed loss. If you only log big breakdowns, performance loss vanishes, yet minor stops and slow running are often the biggest hidden cost.
• Counting rework as good. Anything that needed a second pass is a quality loss.
• A too-slow ideal cycle time. Set performance against the real nameplate rate, not a comfortable target, or you'll mask speed loss.
• Manual logging. End-of-shift clipboards miss the short stops entirely and round the rest. Automatic capture is what makes OEE trustworthy.

Once you have a number, two questions follow: whether it's any good (see what is a good OEE score) and where the losses are hiding (the Six Big Losses).