OEE calculation: the formula, a worked shift, and the six losses.
Overall Equipment Effectiveness is the single number that tells you how much of your planned production time turned into good parts at full speed. It is three simple ratios multiplied together — but most plants compute it wrong by mishandling planned downtime and speed loss. Here is the formula, a full worked example, and the losses behind a low score.
What OEE is
OEE compresses three questions into one percentage: was the machine running when it should have been, did it run at full speed, and did it make good parts? A score of 100 percent would mean only good parts, made as fast as the machine can go, with zero stop time during planned production. No real line hits that, which is why OEE is a tracking and improvement tool rather than a pass-fail gate.
The power of OEE is that it multiplies. A line at 90 percent availability, 90 percent performance and 90 percent quality is not at 90 percent — it is at 0.9 x 0.9 x 0.9 = 72.9 percent. Small losses in three places stack fast.
The three factors
OEE = Availability × Performance × Quality. Each factor is a ratio between 0 and 1:
| Factor | Formula | Captures |
|---|---|---|
| Availability | Run Time / Planned Production Time | Unplanned stops, breakdowns, setup |
| Performance | (Ideal Cycle Time × Total Count) / Run Time | Slow cycles, minor stops |
| Quality | Good Count / Total Count | Scrap, rework, startup rejects |
Two definitions decide everything. Planned Production Time is the shift length minus planned downtime such as scheduled breaks, planned maintenance and booked changeovers. Ideal Cycle Time is the fastest sustainable cycle the machine is designed for, not the average you happen to see.
A worked shift example
Take a single CNC machining cell on an 8-hour shift in a Coimbatore auto-component plant.
| Input | Value |
|---|---|
| Shift length | 480 min |
| Planned breaks | 30 min |
| Planned Production Time | 450 min |
| Unplanned downtime (breakdown + setup) | 47 min |
| Run Time | 403 min |
| Ideal cycle time | 0.5 min/piece |
| Total pieces produced | 760 |
| Reject pieces | 18 |
| Good pieces | 742 |
Now the three factors:
- Availability = 403 / 450 = 0.896 (89.6%)
- Performance = (0.5 × 760) / 403 = 380 / 403 = 0.943 (94.3%)
- Quality = 742 / 760 = 0.976 (97.6%)
OEE = 0.896 × 0.943 × 0.976 = 0.825, or 82.5 percent. That cell is close to world class. The biggest single lever here is availability: the 47 minutes of unplanned stops cost more OEE than speed and quality losses combined.
What counts as a good score
| OEE | Verdict |
|---|---|
| 100% | Theoretical perfection — only good parts, full speed, no stops |
| 85% | World-class for discrete manufacturing (90% A × 95% P × 99% Q) |
| 60% | Typical for a plant that has just started measuring |
| 40% | Common but low — large, recoverable losses present |
Do not chase the 85 percent figure blindly. A line making low-margin parts may rationally run below it; a bottleneck constraint should be pushed hard. OEE is most useful as a trend on the same line over time, not as a league table across different machines.
The six big losses
Every point of lost OEE maps to one of six losses, grouped under the three factors:
- Breakdowns (availability) — equipment failure, the classic unplanned stop.
- Setup and adjustment (availability) — changeover and warm-up time; target with SMED.
- Idling and minor stops (performance) — jams, misfeeds, brief halts under five minutes that rarely get logged.
- Reduced speed (performance) — running below ideal cycle time because of wear, tooling or caution.
- Startup rejects (quality) — scrap made while the process stabilises after a start.
- Production rejects (quality) — scrap and rework during steady-state running.
Common mistakes
- Counting planned downtime against availability. Breaks and scheduled maintenance come out of planned production time first. Including them double-penalises the line.
- Using average cycle time as ideal. Performance must be measured against the fastest sustainable cycle, or speed loss disappears from the number.
- Excluding rework from rejects. Quality should count only first-pass good parts; reworked pieces are a loss even if they eventually ship.
- Comparing OEE across different machines. Different ideal cycle times make cross-machine comparison meaningless. Trend each asset against itself.
- Gaming the number. Inflating planned downtime to lift availability hides the very losses OEE exists to expose.
Connecting OEE to quality data
The quality factor in OEE is only as good as your reject counting, and reject counting depends on disciplined inspection. If your scrap and rework numbers come from a clean, ballooned inspection process, your quality factor is trustworthy. CadNexa's auto-ballooning tool turns a PDF drawing into a numbered inspection sheet, so every defect maps to a specific characteristic — which makes the quality losses in your OEE traceable rather than a guess.
For the wider picture, pair OEE with Cp and Cpk to judge process capability, and use the takt time calculator to check whether your run rate actually meets demand. Standard production-tracking forms are on the templates page.