Illustrative sewing shift: availability 87.5 percent, performance 95.2 percent, quality 95.0 percent; 380 good garments.
GFT instruments / Production performance
Find the loss.
Before adding a line.
Model a sewing shift, reconcile six losses, compare scenarios and export the assumptions behind availability, performance and first-pass quality.
OEE 79.2%60 min stopped380 of 480 min good
An instrument for the decision
How much of the line's time turns into good garments?
Illustrative presets
Every preset is illustrative. No GFT programme OEE default has been supplied.
0The shift, laid straight480 min planned
- Stopped
- Speed loss
- Rejected time
- Good time
Availability87.5%
Performance95.2%
Quality95.0%
One line. One product. One time boundary.
Schematic, not a station simulation. Garment movement is qualitative; the counts and loss minutes carry the calculation. The drawing does not simulate operators, queues, work in progress, machine reliability or station-level bottlenecks.
Good time / Planned time
- Availability
- Performance
- Quality
87.5% × 95.2% × 95.0%
Good output at ideal rate
Availability is the largest loss: 60.0 minutes. Review the stop log before testing a recovery action.
Illustrative worked shift
Live arithmetic, not an operating forecast. A higher score alone does not prove a better investment.
The six big losses
Every lost minute has a bucket.
Each bobbin is wound to its minutes on one shared scale, under the loop it was drawn from.
Full bobbin = 50 min, 10.4% of planned time
Availability
Performance
Quality
The default allocation divides each parent loss equally between its two causes solely to demonstrate the six-loss view. Equal splits are illustrative attribution, not measured causes.
60.0 + 20.0 + 20.0 + 380.0 = 480.0 min. The whole shift reconciles.
Method / Reproducible arithmetic
The time identity checks the score.
1 / 5 · Boundary
- Planned time T
- Stopped, D
- Speed loss
- Rejected time
- Good time
0Ledger, minutes480
T planned minutes, D downtime minutes, c ideal seconds per garment, N total garments, R first-pass rejected garments. 60.0 + 20.0 + 20.0 + 380.0 = 480.0 min. The whole shift reconciles.
Step 1 / 5 · Boundary
Start with one production boundary.
Use a single line and a single product for the shift. Planned time includes only time when production was intended. Exclude breaks when the line was never scheduled to run. Write that convention into the supervisor's shift record before comparing days.
Running minutes = T − DStep 2 / 5 · Stops
Count every stop once.
Count changeovers inside the plan as downtime. Keep minor stops inside running time when they are assigned to performance, otherwise the same stop would be counted twice. Availability is the running share of the plan.
Availability = (T − D) / TStep 3 / 5 · Rate
Set the rate against an ideal cycle.
The cycle input is the ideal elapsed line cycle per completed garment, in seconds. A garment's standard minute value (SMV) measures labour content and cannot be inserted here without converting it into an effective line rate. Enter a validated line pitch, or keep the result illustrative.
Performance = c × N / [60 × (T − D)]Step 4 / 5 · Rejects
Take out the first-pass rejects.
Quality is good garments over total garments. For an empty shift, undefined performance or quality factors are displayed as zero by convention. No completed garments means no good-equivalent time, and a fully stopped shift cannot also report production.
Quality = (N − R) / NStep 5 / 5 · Reconcile
The four parts must add back to the plan.
The worked shift converts 480 planned minutes into 380 good-equivalent minutes: 60 stopped, plus 20 speed loss, plus 20 reject-equivalent, plus 380 good, is 480 in all. The inputs are illustrative, chosen to make the arithmetic inspectable. It is not a finding about a factory.
OEE = A × P × Q = 380 / 480
Interpretation / Conditions
A score identifies a question to investigate.
- Fork
- Trade
- Not utilisation
Investigate the largest loss first.
The three parent losses are ranked in minutes. They do not infer a maintenance diagnosis: reason-coded stop events and quality records are needed to replace assumed attribution with measured evidence.
Moving minutes between stopped and running time does not move OEE.
With the good count held fixed, availability and performance move in opposite directions. A genuine recovery needs a justified change in output, quality or the production boundary.
Do not substitute utilisation for OEE.
The programme annex records CICAM capacity utilisation from 10.83 percent at CICAM 2 to 48.9 percent at CICAM 1 in a July 2023 report. This is historical context from R12, not an OEE measurement, a current operating assessment or a target for this simulator.
Tap a loop to read its question
Availability. Review the stop log before testing a recovery action. Return to the console ↑
Performance. Check small stops and line balancing against a validated ideal cycle. Open line balancing ↗
Quality. Review first-pass defects and startup rejects before accepting more output. Open inspection economics ↗
Availability 87.5% · Performance 95.2% · OEE holds at 79.2%
38.07 points apart (48.9 minus 10.83): utilisation, not OEE.
A / B decision record
Compare the change, keep the boundary.
Saved A starts as the illustrative stop-heavy preset, so the comparison can be read at once. Same planned time and ideal line cycle. Delta is current B minus saved A; allocation changes alone do not change OEE.
No scenario A is saved. Set the shift, then choose Save current as A. Edit the inputs again to form B.
59.4%OEE
79.2%OEE
B minus A+19.8 pp
0Both threads on one scale480 min
Stopped: A 120.0 min · B 60.0 min · -60.0 min
| Measure | Saved A | Current B | Delta |
|---|---|---|---|
| OEE | 59.4% | 79.2% | +19.8 pp |
| Good garments | 285 | 380 | +95 |
| Good time (min) | 285.0 | 380.0 | +95.0 |
| Planned time (min) | 480.0 | 480.0 | 0.0 |
R12 qualifiedProgramme-annex utilisation, not OEE; not independently reverified here.
Assumptions and sources / Status register
Defaults illustrate a method, not a benchmark.
Default shift: 60 min stopped, 400 completed, 20 rejected. Illustrative, chosen to make the arithmetic inspectable.
| Input | Starting value | Live value | A | P | Q | Status |
|---|---|---|---|---|---|---|
| Planned time | 480 min | Does not move quality | Illustrative | |||
| Downtime | 60 min | Does not move quality | Illustrative | |||
| Ideal line cycle | 60 sec/garment | Does not move availability | Does not move quality | Illustrative | ||
| Total completed | 400 garments | Does not move availability | Illustrative | |||
| First-pass rejected | 20 garments | Does not move availability | Does not move performance | Illustrative | ||
| Cause allocations | 50% each | Does not move availability | Does not move performance | Does not move quality | Allocation |
- Sources. The equations follow Vorne's OEE calculation method; the cause categories follow its Six Big Losses framework. Both pages were opened for this implementation on 30 October 2026; their publication dates were not displayed. Source receipts are retained with this tool. R12 remains qualified programme-annex evidence and was not independently reverified here.
- Sharing. Saved A is held in this page's memory. A shared link includes A and current B, and restores their inputs rather than a rounded score. CSV exports include raw inputs, unrounded outputs, units, allocation status and above-ideal diagnostics. Sharing copies a link locally; it does not submit shift records to GFT. Review a link before distributing commercially sensitive production information.
- Limits. A planning aid. Verify inputs and boundaries before an operational decision. The model does not replace measured shift records, a validated cycle standard and engineering review.
WP02 Measure Before You Model, WP15 Utilisation Decides, WP18 Edge to Dashboard sit behind this instrument. Related papers in the research programme; publication status is controlled by the library. Research library →
Use the result in its boundary
A shift is a scenario, not a forecast.
The example does not predict a factory's performance. Replace it with measured inputs from the same line, product and time window before drawing an operating conclusion.
Minute 30 of 480 · stopped time, 60.0 min (Availability) · Which stops were unplanned breakdowns, and how long did each last?
At B: your current shift. Turn the bezel toward A to replay the saved scenario.
The model stays visible: availability × performance × quality, with planned production time as the denominator. A planning aid. Not investment, legal or tax advice. Verify inputs and boundaries before an operational decision.
87.5% × 95.2% × 95.0% = 380 ÷ 480 = 79.2%
Laid in ledger order, not in the order events happened