Overall Equipment Effectiveness (OEE)

OEE multiplies three losses that used to be reported separately: availability (did it run when it was supposed to), performance (did it run at rate), and quality (was the output good). Seiichi Nakajima packaged them in the 1980s as the scoreboard for total productive maintenance, and the multiplication is the whole point. Ninety percent availability, ninety-five percent performance and ninety-nine percent quality look like three excellent numbers, and $0.90 \ imes 0.95 \ imes 0.99 = 0.846$. That 85% figure is the widely quoted world-class benchmark, and most plants that measure honestly for the first time land between 40% and 60%.

The classic mistake is averaging the three instead of multiplying them. The same three figures average to 94.7%, which is off by nine percentage points and always flattering. The other classic mistake is gaming the availability denominator. If unstaffed shifts, planned maintenance and setup time are all quietly excluded from "planned production time", OEE climbs without anything improving. Two plants cannot compare OEE unless they agree on that denominator first, and they usually do not.

Read it as a diagnostic rather than a grade. A single OEE number tells you almost nothing; the three components tell you where to go. Low availability is a maintenance and changeover problem, low performance is usually minor stoppages and reduced speed, and low quality is a process control problem. Tracking the product alone hides which one moved.