The machine ran all day and the order is still short. Overall Equipment Effectiveness explains why, by showing how many of your scheduled hours turned into good product.
Most plants know how many units they shipped. Far fewer know how many they could have shipped with the same machine, the same people and the same shifts. That gap is hidden capacity, and you are already paying for it.
Overall Equipment Effectiveness (OEE) measures the gap. It is one percentage, built from three questions about a piece of equipment.
Three questions, one number
| Factor | The question | How to calculate it |
|---|---|---|
| Availability | Was the machine running when it was scheduled to run? | Operating time ÷ scheduled time |
| Performance | When it ran, did it run at its ideal speed? | (Ideal cycle time × total units) ÷ operating time |
| Quality | Of the units it made, how many were good the first time? | Good units ÷ total units |
OEE = Availability × Performance × Quality
Because the three factors multiply, OEE is also the share of scheduled time that produced good units at ideal speed. Everything else is a loss.
A worked example: one day on a packaging line
Take a line that runs two 8.5-hour shifts. Breaks and a shift meeting take 2 hours, so the line is scheduled for 15 hours. The machine is designed for one unit every 1.8 seconds, which is 2,000 units per hour.
On this day the line lost 2.25 hours to a breakdown and two changeovers. While it ran, it averaged 1,800 units per hour. It made 22,950 units and 918 of them were rejected.
Table 1. The OEE calculation for the example day
| Step | Calculation | Result |
|---|---|---|
| Scheduled time | 17.0 h total − 2.0 h planned stops | 15.00 h |
| Operating time | 15.0 h − 2.25 h downtime | 12.75 h |
| Availability | 12.75 ÷ 15.00 | 85.0% |
| Performance | 22,950 units ÷ (2,000 units/h × 12.75 h) | 90.0% |
| Quality | 22,032 good ÷ 22,950 total | 96.0% |
| OEE | 85.0% × 90.0% × 96.0% | 73.4% |
Each factor looks respectable on its own. Together they mean that only 11 of the 15 scheduled hours produced sellable product. The other 4 hours were paid for and lost.
Try it: move the losses and watch OEE
Fixed for this example: 17 hours of total time, 2 hours of planned stops and an ideal rate of 2,000 units per hour.
The six big losses
Every lost hour falls into one of six categories. Naming them matters because each one has a different owner and a different fix.
Table 2. The six big losses and the factor each one reduces
| Factor | Loss | What it looks like on the floor |
|---|---|---|
| Availability | 1. Equipment failure | Breakdowns, waiting for maintenance or parts |
| 2. Setup and adjustment | Changeovers, warm-up, material shortages | |
| Performance | 3. Idling and minor stops | Jams, blocked sensors, stops under a few minutes that nobody logs |
| 4. Reduced speed | Running below the ideal rate because of wear, material or habit | |
| Quality | 5. Process defects | Scrap and rework during steady production |
| 6. Startup rejects | Scrap made after a start or changeover until the process settles |
Four things that distort the number
- Planned stops are outside OEE. Breaks and planned maintenance are removed before the calculation starts. If you measure against all 17 hours instead, you get TEEP (Total Effective Equipment Performance), which is 64.8% here. Use OEE to judge how well the line runs. Use TEEP to judge whether you need more equipment or simply more scheduled hours.
- The ideal cycle time must be the true best rate. If the standard is set at a comfortable speed, performance shows 100% and the speed loss disappears from view. Use the design rate or the best rate the line has sustained.
- Three good scores multiply into a mediocre one. A line at 90% on every factor has an OEE of 72.9%. This is why a plant can feel busy and still fall short.
- OEE is for machine-paced work. It suits automatic and semi-automatic equipment. Manual assembly has no fixed ideal cycle time, so labour productivity measures fit better there.
A figure of 85% is often quoted as world class. Treat it as a reference point only. A line with frequent changeovers will never match a dedicated line, so the useful comparison is the same line against itself over time.
Takeaway. Before any improvement project, measure where the hours go for a few weeks. The 73.4% matters less than the split behind it: 2.25 hours of downtime, 1.3 hours of speed loss and half an hour of rejects.
The line, its revenue and the project in this series are an illustrative example. The calculators are screening tools and do not replace a full financial analysis.
