Blog · How-to guides · Manufacturing
The six big losses separate downtime, slow production and rejected output. Use recorded production exports to compare losses by line and reconcile them to planned production time.
The six big losses in total productive maintenance explain why scheduled production does not become good output at the ideal production rate. They separate equipment failure, setup and adjustments, minor stops, reduced speed, process defects and startup rejects. Use them to choose a specific improvement on a specific line, with recorded minutes and units behind the choice.
The categories connect maintenance work to production results. OEE.com's six-loss definitions group the first pair under availability, the next pair under performance and the last pair under quality.
| Loss | What belongs in it | First owner to investigate |
|---|---|---|
| Equipment failure | Significant stops caused by breakdowns or other agreed unplanned reasons | Maintenance lead with production |
| Setup and adjustments | Changeovers and adjustments during scheduled production | Production lead with setup team |
| Idling and minor stops | Brief interruptions recorded below the significant-stop threshold | Line supervisor |
| Reduced speed | Running below the agreed ideal production rate | Process engineer with operators |
| Process defects | First-pass rejects during stable production | Quality lead |
| Startup rejects, also called reduced yield | First-pass rejects before stable production | Setup team with quality |
Assign an event to one category. Agree a stop-duration boundary locally, and apply it consistently. A jam recorded as significant downtime cannot also be included in minor-stop minutes.
Keep this analysis narrower than a general manufacturing KPI report. Its purpose is to account for lost productive time, rather than score every part of the plant.
Use a US packaging manufacturer's shift records. Both lines make the same item, with an ideal cycle time of 0.5 minute per unit. Planned production time already excludes periods when the lines were not scheduled to run.
The export should identify the date, shift, line, item, planned minutes, significant-stop reasons, minor-stop minutes, total units and first-pass reject reasons. Keep the cycle-time reference with the report so a later revision does not silently change history.
| Recorded input | Line A | Line B |
|---|---|---|
| Planned production time, minutes | 420 | 420 |
| Equipment failure, minutes | 40 | 15 |
| Setup and adjustments, minutes | 20 | 30 |
| Minor stops, minutes | 25 | 20 |
| Ideal cycle time, minutes per unit | 0.5 | 0.5 |
| Total units produced | 600 | 650 |
| Startup rejects, units | 20 | 10 |
| Process defects, units | 10 | 20 |
These are illustrative records, not an industry benchmark. If a line runs different items, calculate ideal production minutes for each item before adding them together.
Downtime is already measured in minutes. Convert quality losses to equivalent ideal-production minutes, which lets all categories reconcile to one denominator. These equivalent minutes describe lost output capacity; they are not extra clock time spent inspecting rejected units.
Run time = Planned production time − Equipment failure − Setup and adjustments
Ideal time for total output = Total units × Ideal cycle time
Reduced-speed loss = Run time − Ideal time for total output − Minor-stop minutes
Quality-loss minutes = Rejected units × Ideal cycle time
Line A has 420 − 40 − 20 = 360 run minutes. Its recorded output represents 600 × 0.5 = 300 ideal minutes. The performance gap is therefore 360 − 300 = 60 minutes, comprising 25 minor-stop minutes and 35 reduced-speed minutes.
Line B has 420 − 15 − 30 = 375 run minutes. Its output represents 650 × 0.5 = 325 ideal minutes, leaving 50 performance-loss minutes: 20 minor-stop minutes and 30 reduced-speed minutes.
| Loss, in equivalent production minutes | Line A | Line B |
|---|---|---|
| Equipment failure | 40 | 15 |
| Setup and adjustments | 20 | 30 |
| Minor stops | 25 | 20 |
| Reduced speed | 35 | 30 |
| Process defects | 10 × 0.5 = 5 | 20 × 0.5 = 10 |
| Startup rejects | 20 × 0.5 = 10 | 10 × 0.5 = 5 |
| Total losses | 135 | 110 |
The residual calculation identifies the size of reduced-speed loss; it does not establish the cause. Review the underlying shift notes before attributing the gap to tooling, materials or operator practice. A downtime dashboard can support a separate review of significant-stop reasons.
Line A produces 600 − 20 − 10 = 570 good units, equivalent to 570 × 0.5 = 285 productive minutes. Line B produces 650 − 10 − 20 = 620 good units, equivalent to 310 productive minutes.
| Reconciliation | Line A | Line B |
|---|---|---|
| Productive minutes | 285 | 310 |
| Loss minutes | 135 | 110 |
| Productive minutes + losses | 420 | 420 |
| Planned production minutes | 420 | 420 |
| Unexplained difference | 0 | 0 |
OEE.com's calculation guide gives the equivalent identity: OEE equals good units multiplied by ideal cycle time, divided by planned production time. Here that is 285 / 420 = 67.9% for A and 310 / 420 = 73.8% for B, rounded to one decimal place.
Across both lines, productive minutes are 285 + 310 = 595 and planned minutes are 420 + 420 = 840. Combined OEE is 595 / 840 = 70.8%. Use summed productive and planned minutes for larger groups, especially when shift lengths differ.
Line A's largest individual loss is equipment failure. Maintenance should review the recorded failures with production, select a documented cause and check the same category after the corrective work. Line B's setup and reduced-speed losses are tied, so the setup lead and process engineer should inspect those records before choosing a project.
Plant management owns the common definitions. Production owns stop coding, quality owns reject classification, and maintenance owns repair evidence. Finance can connect an approved improvement to manufacturing cost per unit, but equivalent production minutes are not automatically cash savings. Recovered capacity only has financial value under stated labor, demand and cost assumptions.
Bring a production export with line, shift, planned minutes, recorded stops, output and reject counts. Covirage maps the columns, and its tools compute and check the loss comparisons; the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own export.
See Covirage for industrial manufacturers for the business context.
Equipment failure and setup reduce availability. Minor stops and reduced speed reduce performance. Process defects and startup rejects reduce quality. Overall equipment effectiveness combines those three factors, while the six categories show which recorded losses sit behind the combined result.
A changeover inside time scheduled for production is an availability loss in this calculation. Time when production was never scheduled sits outside the denominator. Document that boundary before comparing lines so that moving a changeover between categories cannot create an apparent improvement.
For a first-pass quality measure, units needing rework are rejected on their original pass. Later recovery does not erase the original quality loss. Keep rework movements separate so the same physical unit is not counted again as new production.
Yes, when the records contain planned time, significant stops, output and rejects, plus an agreed ideal cycle time. Separating minor stops from reduced speed also needs a recorded minor-stop total. Without it, report a combined performance loss rather than inventing the split.