TL;DR. Every plant loses time in the same six ways. They are easy to name and hard to see, because they hide in different records and happen on different shifts.
The six, in three groups
| Group | Loss | What it looks like on the floor | |---|---|---| | Availability | 1. Equipment failure | The line stops and stays stopped until repaired, usually five minutes or more. | | Availability | 2. Setup and changeover | Switching product, size or recipe, then adjusting until it runs right. Only the time beyond the agreed standard counts as the loss. | | Performance | 3. Minor stops and idling | Jams, sensor trips and short waits. Each is usually under about five minutes and rarely logged. | | Performance | 4. Reduced speed | The line runs, but below its rated speed. | | Quality | 5. Process defects | Scrap and rework made while running normally. | | Quality | 6. Start-up rejects | Waste while the line warms up or settles after a restart. Required first-off checks are not a loss, only the avoidable waste is. |
One line, one shift
Illustrative example. The numbers are made up to show the shape, not a measured result.
A packaging line has 480 minutes of planned production time. Standard changeover time is treated as planned and is not in the 480.
| Where the minutes go | Minutes | |---|---| | Planned production time | 480 | | A drive fault stops the line (loss 1) | -30 | | Two changeovers that overran the standard time (loss 2) | -40 | | Dozens of short jams and waits (loss 3) | -25 | | Running below rated speed, expressed as time (loss 4) | -20 | | Rework and rejects, expressed as time (losses 5 and 6) | -15 | | Fully productive time | 350 |
That is about 73% of planned time. Speed and quality losses are expressed as the time they cost, so 350 out of 480 is the same figure OEE reports. On paper the line ran all shift.
Why the losses hide
- Different owners. Maintenance sees failures. Production sees changeovers. Quality sees scrap. Nobody sees the total.
- Different records. Stops live in the control system, reasons in a log, scrap in a quality sheet.
- Small and frequent. A two-minute stop thirty times is an hour, and no single stop looked worth a note.
- Averages. A monthly average hides the one product, shift or machine that causes most of the loss.
What to do with this
- Count minutes lost in each of the six groups for one line over one month. A rough count beats none.
- Find which group is biggest. It is rarely the one people assume.
- Pick one recurring loss in that group and give it a single owner and a date.
- Check again next month on similar production, not just the calendar.
Common questions
Is every changeover a loss? No. The required part is planned time. The overrun beyond the agreed standard is the loss.
Do all six apply to every plant? In some form, yes. A batch process shows them as holds and rework, a discrete line as stops and scrap.
Where does the list come from? From total productive maintenance, and it is the usual starting point for OEE. See the OEE overview.