A short crew shows up in all three OEE factors: stops with nobody to cover them hit availability, a slower line hits performance, and rushed or new staff hit quality. Keep the ideal rate the same, code the labor reason, and if you decided before the shift not to run a line, take it out of planned time and report those hours apart.
Why a short crew is easy to hide
When people are missing, the natural reaction is to adjust the expected rate, since nobody can blame the line for running at 4 people's pace. That feels fair, and it makes the problem disappear from the report.
A missing person does not change what the line can do, so the shortage should show up as a loss.
OEE measures the line against that capability, as availability x performance x quality, the method in Overall Equipment Effectiveness (R. C. Hansen, 2001).
Where it shows up
| Factor | What you see on the floor | How to code it |
|---|---|---|
| Availability | Station empty, line stopped. Breaks stop the line because nobody can relieve | Labor: no cover |
| Performance | Speed turned down to match the people. Short waits nobody logs | Labor: reduced speed |
| Quality | More scrap from new or stretched staff. Checks missed | Reason code plus crew note |
Logging who was missing, and how to split the reasons, is covered in waiting for operator downtime. This page is about the whole shift's numbers, and the decisions they should drive.
Worked example: the same line, one person short
Example: a line with a standard crew of 5, an ideal rate of 40 units a minute and 450 planned minutes. Today one person is absent, so the lead turns the line down to about 30 a minute and loses extra time to uncovered breaks.
| Crew 5 | Crew 4, ideal 40/min | Crew 4, ideal lowered to 30/min | |
|---|---|---|---|
| Run time | 390 min | 355 min | 355 min |
| Total units | 14,040 | 9,900 | 9,900 |
| Good units | 13,759 | 9,455 | 9,455 |
| Availability | 86.7% | 78.9% | 78.9% |
| Performance | 90.0% | 69.7% | 93.0% |
| Quality | 98.0% | 95.5% | 95.5% |
| OEE | 76.4% | 52.5% | 70.0% |
The line made 31% fewer good units, and the honest figure falls by 24 points. The last column hides most of that fall by moving the goalposts.
That lowered rate also ruins every later comparison, since last month's OEE and this month's now use different yardsticks. See ideal cycle time for why one rate per product must stay fixed.
Count the loss in minutes
Good units divided by the ideal rate gives good minutes. That turns the shortage into a figure you can put next to the other losses.
Compare those 108 minutes with what it costs to fix them, such as overtime, agency cover or moving someone from another line. Then the downtime cost calculator can put a value on it.
The labor view: same shift, different story
OEE and labor numbers can point in different directions, so look at both. Units per labor hour on a 7.5-hour shift is 13,759 / 37.5 = 367 with the full crew, and 9,455 / 30 = 315 with one person short.
So here, a crew 20% smaller made 31% fewer units. This line needs its fifth person, since each hour with 4 people yields less per person, and only a test like this can show you that.
On another line the result could be the opposite, with one less person costing little. Overall labor effectiveness shows the same event from the people's side.
The counting rules
Most arguments disappear if you decide these in advance. The test is when the decision was made, and whether the line was on the schedule.
Stays in planned time. The losses count in all three factors at the normal ideal rate.
Out of planned time, since it was not scheduled. Log the hours as no-crew hours, because OEE alone will not show them.
Availability loss for the rest of the planned time, coded as labor.
The first line logs labor loss, or is rescheduled if the decision came first. The other line is measured as usual.
Rule 2 is the honest gap in OEE. A line taken off the plan does not hurt OEE at all, so a plant could cancel half its shifts and still post a fine number.
That is why you also need a wider measure. TEEP and utilization both show the hours OEE leaves out, and OEE with no orders handles the similar case when demand is missing.
What to do with the numbers
- Build a crew table per line. For each crew size, test and write down the rate the line can really hold. Use it for planning, while OEE stays at the ideal rate.
- Staff the constraint first. With 8 people and two 5-person lines, cover the bottleneck fully before you split them evenly.
- Look for jobs that can wait. Some roles can be shared across lines on a short day, and a gemba walk shows which ones.
- Cross-train before you need it. A skills matrix tells you who can step in, and OEE training for operators helps them log the event.
- Change the plan early. A plan that assumes a full crew makes plan attainment fail all week, so plan to the crew you really have.
Common mistakes
- Lowering the ideal rate for the day. Report the crew-adjusted rate as a separate planning figure, if you need one.
- Coding it as a machine problem. A slow line gets blamed on the machine, and engineering chases the wrong issue.
- Not noting crew size. Without crew size per shift you cannot learn what each person is worth, so add it to the shift handover.
- Calling new-staff scrap a quality issue only. Tag it with the crew note, so you can see the training effect.
A short crew shows first as slow running and short waits, the losses a paper log tends to miss. Fabrico records stops, speed and counts live from the line, and operators tag the cause at the line, so a labor reason can sit next to the time it cost. See how Fabrico logs stops and causes.
What people ask next
Does a labor shortage count against OEE?
Yes, if the line is scheduled to run. Stops with no cover hit availability, reduced speed hits performance and extra scrap hits quality, all at the normal ideal rate.
Should I lower the ideal cycle time when the line is short-staffed?
No. Keep the ideal cycle time fixed, or the shortage disappears from OEE and your trend stops being comparable. Use a separate crew-based rate for planning.
What if we decide not to run a line because we have no crew?
If the decision comes before the shift, the line is not scheduled, so those hours leave planned time. Log them as no-crew hours, and watch TEEP or utilization so they stay visible.
How do I measure the cost of being one person short?
Compare good minutes on full-crew and short-crew shifts of the same product, then value the gap. The downtime cost calculator helps with the last step.
Is overall labor effectiveness better than OEE for this?
They answer different questions, so use both. OEE shows what the line lost, and overall labor effectiveness shows how well the people's time was used.
Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.
Related guides: Cost Per Unit: Formula and How OEE Changes It · OEE for Assembly Lines: Takt and Station Data · OEE for Manual Processes: How to Measure It · How to Calculate Line Efficiency: Both Formulas