A machine that is “always running”
A production manager is told that the new machining centre is fully utilised: it runs on every shift. Yet the orders are always late. Looking closer, the machine stops several times a shift for material, runs below its rated speed because of a vibration problem, and some parts are scrapped at the end of the line. Each loss seems small. Together they explain why a machine that is always “on” delivers far less than it could.
OEE puts a number on that gap.
The three factors
OEE was introduced by Seiichi Nakajima as part of Total Productive Maintenance (TPM), and is defined in ISO 22400-2:2014 as the product of three ratios:
OEE = availability × performance × quality
- Availability: the share of the planned production time during which the machine actually runs. Stops reduce it.
- Performance: the actual output compared with what the machine could have made in the time it ran, at its ideal cycle time. Slow cycles and minor stops reduce it. ISO 22400-2 calls this factor effectiveness.
- Quality: the share of produced parts that are good, without scrap or rework.
The losses are often grouped into Nakajima’s “six big losses”: breakdowns and set-ups or adjustments (availability), idling, minor stops and reduced speed (performance), and defects and start-up losses (quality).
Worked example
Illustrative values for one shift:
- Planned production time: 450 minutes. Stops for breakdowns, material and set-up: 81 minutes. Run time: 369 minutes.
- Availability = 369 / 450 = 0.82
- Ideal cycle time: 30 seconds, so in 369 minutes the machine could make 738 parts. It made 627.
- Performance = 627 / 738 = 0.85
- Of the 627 parts, 577 were good.
- Quality = 577 / 627 = 0.92
OEE = 0.82 × 0.85 × 0.92 ≈ 0.64, or 64%.
A useful check: good parts × ideal cycle time / planned time = 577 × 0.5 min / 450 min ≈ 0.64. OEE is simply the share of the planned time that would have been needed to make the good parts at full speed.
Reading OEE
The single OEE number says how big the gap is. The three factors say where it is. In the example, availability is the largest loss: 81 minutes of stops. That is where to look first, and production monitoring with stop causes shows which stops matter most.
Because the factors multiply, three numbers that each look reasonable, around 80 to 90%, give an OEE of only around 60%. That is normal and not a sign that something is wrong with the calculation.
Getting the definitions right
OEE figures are only comparable if they are calculated the same way. Decide and document:
- What counts as planned time. Are breaks, planned maintenance and shifts without orders excluded? ISO 22400-2 distinguishes between planned busy time and other time categories.
- Which cycle time is “ideal”. The rated speed of the machine, the best demonstrated speed, or the standard time in the routing? A generous ideal cycle time inflates performance.
- What counts as good. Are reworked parts good, or only those right first time?
Common mistakes
- Comparing machines or plants on OEE. Different products, batch sizes and definitions make such comparisons misleading. OEE is best used to follow one machine or line over time.
- Chasing the number. Reclassifying stops as planned time raises OEE without improving anything.
- Ignoring quality. A machine that runs fast but makes scrap may have high availability and performance, but poor quality pulls OEE down, as it should. The causes of scrap belong to quality assurance as much as to maintenance.
- Relying on manual records. Short stops are rarely written down by hand, so manually recorded OEE tends to be too high. Automatic collection from the machine, as part of Industry 4.0, usually reveals more losses.
In short
OEE shows how much of the planned time became good parts, and the three factors show whether the time was lost to stops, slow running or defects. Use it to find the biggest loss on a machine, fix it, and follow the result over time.
