OEE explained

Overall Equipment Effectiveness (OEE) measures how much of a machine's planned production time went to making good parts at full speed. This guide works through one example shift.

Matt Savas · Reviewed October 2026 · About 5 minutes

What OEE measures

OEE comes from Total Productive Maintenance (TPM), the maintenance system Seiichi Nakajima developed at the Japan Institute of Plant Maintenance in the 1970s and 1980s. It combines three measures into one percentage.

  • Availability is the share of planned time the machine was running.
  • Performance is how fast it ran compared with its ideal speed.
  • Quality is the share of parts that were good the first time.

OEE is the three multiplied together.

Why plants measure it

Most plants know when a machine was down for an hour. Fewer know about the 35 minutes of short stops, or the eight seconds lost on every cycle. OEE puts all of these losses on one scale, minutes of planned time, so you can see which are largest and where to start.

How to calculate OEE

  1. Take the shift length and subtract breaks, meetings and time with no orders. The result is planned production time.
  2. Availability is run time divided by planned production time.
  3. Performance is ideal cycle time × parts made, divided by run time.
  4. Quality is good parts divided by all parts made.
  5. OEE is availability × performance × quality.

A worked example

One machine runs a 480-minute shift. Its ideal cycle time is 60 seconds, so at full speed it makes one part a minute. Breaks and a period with no orders take 50 minutes, which leaves 430 minutes of planned production time.

Breakdowns and changeovers take 75 minutes, so the machine runs for 355 minutes. Availability is 355 ÷ 430 = 82.6%. During those 355 minutes it has 35 minutes of short stops and runs at 68 seconds a part instead of 60. It makes 282 parts when it could have made 355, so performance is 79.4%.

Of the 282 parts, 23 are defects or startup rejects after a changeover. Quality is 259 ÷ 282 = 91.8%. OEE is 82.6% × 79.4% × 91.8% = 60.2%, which is the same as 259 good parts in 430 planned minutes.

82.6%
Availability
79.4%
Performance
91.8%
Quality
60.2%
OEE

Change the example

Move the sliders to change the shift. The bar and the four numbers update as you go.

One 480-minute shift
OEE 60.2%
Good parts259
Quality loss23
Performance loss73
Availability loss75
Planned downtime50

All values in minutes.

Availability
82.6%
Performance
79.4%
Quality
91.8%
TEEP
54.0%

Why the factors are multiplied

Each factor only measures the time the one before it left. Performance looks only at run time, and quality looks only at the parts that were made. Multiplying them gives the share of planned time that produced good parts.

It also means that modest losses add up. A machine at 90% availability, 90% performance and 90% quality has an OEE of 73%. An average of the three would say 90%.

The six big losses

Nakajima grouped the losses behind OEE into six categories, two for each factor. They make a useful checklist when one factor comes in low.

FactorLossExamples
AvailabilityBreakdownsEquipment failures during planned time
ChangeoversSetup and adjustment between products
PerformanceMinor stopsJams, misfeeds and sensor trips
Reduced speedRunning slower than the ideal cycle time
QualityDefectsParts scrapped or reworked
Startup rejectsScrap after a start or changeover

Common mistakes

  • Comparing different machines. OEE depends on the product mix and the number of changeovers, so a line that runs one product all week will score higher than a cell that changes over six times a day. Compare each machine with its own history.
  • Improving machines that aren't the bottleneck. Raising OEE on a machine that isn't the constraint doesn't raise the line's output. It usually builds inventory in front of the constraint.
  • Changing the ideal cycle time. In the example, changing the ideal from 60 to 70 seconds raises performance by 13 points without one extra part. Use the rate the equipment can actually sustain, and keep it fixed.
  • Relying on the machine's counter. Controllers usually record long breakdowns but not short stops. A manual study often finds that minor stops are the largest loss.

OEE and TEEP

Total Effective Equipment Performance (TEEP) uses the same good production time but divides it by all the time in the period, including breaks and time with no orders. In the example, TEEP is 259 ÷ 480 = 54.0%. Plants use OEE to find and reduce losses, and TEEP to see how much spare capacity they have before buying more equipment.

How to start

Pick one machine, ideally the bottleneck. For a few shifts, have someone record every stop with the time and the reason, count parts and rejects, and time the actual cycle with a stopwatch. Most first measurements come in lower than people expect, and the record of stops shows which losses to work on first.

When you have the numbers, the free OEE calculator does the arithmetic. Changeovers are often the largest loss; SMED is the standard method for shortening them.

Frequently asked questions

What is OEE and how is it calculated?
OEE measures how much of a machine's planned production time was spent making good parts at full speed. It is availability × performance × quality. Availability is run time ÷ planned production time. Performance is ideal cycle time × parts made ÷ run time. Quality is good parts ÷ parts made. Breaks and periods with no orders are left out of planned production time.
What is a good OEE score?
Figures often cited put a typical plant around 60% and world-class around 85%. They are rules of thumb, not measurements. Because OEE depends on product mix and changeovers, the most useful comparison is a machine against its own history.
What are the six big losses in OEE?
Breakdowns and changeovers reduce availability. Minor stops and reduced speed reduce performance. Defects and startup rejects reduce quality.
What is the difference between OEE and TEEP?
OEE divides good production time by planned production time. TEEP divides it by all the time in the period, so TEEP is never higher than OEE. OEE is used to reduce losses; TEEP shows spare capacity.
Why is OEE multiplied instead of averaged?
Each factor measures only the time the previous one left, so the losses compound. Three factors at 90% give an OEE of about 73%. An average would say 90%.
How do you start measuring OEE?
Start on paper with one machine, ideally the bottleneck. Record every stop for a few shifts, count parts and rejects, time the actual cycle, and calculate the three factors. Automate the data collection later.