OEE differs between shifts for two kinds of reasons. Some are real: product mix, changeovers, staffing, material and maintenance support. Some are measurement: different ideal rates, different scheduled time, different logging habits. Compare the loss breakdown of each shift, not the single OEE number, and check the measurement first.
Do not start with the crew
The first reaction to a gap is "the night shift is weaker". It is often wrong.
A shift with a lower OEE may have a harder product, fewer people, no material at night or a stricter log. Blame comes first, and facts last.
The right order is: check the measurement, then the mix, then the losses, and only then the people.
Step 1: check that you compare like with like
Before you compare, confirm that both shifts use the same rules.
| Check | What can go wrong |
|---|---|
| Same ideal rate per product | One shift uses a lower "ideal" and shows a higher performance |
| Same scheduled time rule | One shift excludes breaks or waiting from scheduled time, the other does not |
| Same stop threshold | One shift logs 30-second stops, the other does not |
| Same code list | Different crews put the same stop under different codes |
| Same product mix | One shift runs the hard product more often |
| Long enough period | A few days can be one odd event |
Overall Equipment Effectiveness (R. C. Hansen, 2001) says to use the same data for everyone, to keep one accredited ideal rate per product family, and to validate the data against hour meters and counts. A gap can shrink or vanish once these checks are done.
Worked example
Example: the same line, ideal rate 60 units per minute, 450 scheduled minutes per shift. Over a month, the average day shift reports 76.2% OEE and the average night shift 69.7%.
| Item | Day | Night |
|---|---|---|
| Runtime | 380 min | 330 min |
| Units made | 21,000 | 19,200 |
| Good units | 20,580 | 18,816 |
| Availability | 84.4% | 73.3% |
| Performance | 92.1% | 97.0% |
| Quality | 98.0% | 98.0% |
| OEE | 76.2% | 69.7% |
Check the day shift: 380 / 450 = 84.4%. 21,000 / (380 x 60) = 92.1%. 20,580 / 21,000 = 98.0%. The product is 76.2%.
The simple route agrees: 20,580 / 60 = 343 minutes, and 343 / 450 = 76.2%.
Check the night shift: 18,816 / 60 = 313.6 minutes, and 313.6 / 450 = 69.7%.
The gap is 6.5 points. But look at performance.
The night shift runs faster when it runs. The loss is all in availability.
Step 2: compare the losses, not the OEE
Put each shift's 450 minutes into loss buckets. The buckets are the ones in the six big losses, grouped as stops, speed and waste.
| Loss | Day min | Night min | Gap |
|---|---|---|---|
| Stops | 70 | 120 | +50 |
| Speed loss | 30 | 10 | -20 |
| Waste | 7.0 | 6.4 | -0.6 |
| Total loss | 107.0 | 136.4 | +29.4 |
| Good time | 343.0 | 313.6 | -29.4 |
Each shift adds up to 450 minutes: 107.0 + 343.0 and 136.4 + 313.6. The night shift loses 50 more minutes to stops and 20.6 fewer to speed and waste. The net gap is 29.4 minutes, which is 6.5 points.
Now drill into the stops with the reason codes.
| Stop cause | Day min | Night min |
|---|---|---|
| Changeover | 30 | 35 |
| Breakdown | 25 | 30 |
| Waiting for material | 15 | 55 |
| Total | 70 | 120 |
Waiting for material accounts for 40 of the 50 extra minutes. Changeover and breakdown are close.
In this example the cause could be that the store is closed at night. The cause is not the crew.
That is a very different conversation from "the night shift is weak". It points to a staging or supply fix. The code list that makes this possible is in downtime reason codes.
Step 3: check the product mix
Mix alone can create a gap. Example: two crews with exactly the same skill. Product X runs at 80% OEE and product Y at 60% for both.
| Shift | X time | Y time | Weighted OEE |
|---|---|---|---|
| A | 300 min | 150 min | (300 x 80% + 150 x 60%) / 450 = 73.3% |
| B | 150 min | 300 min | (150 x 80% + 300 x 60%) / 450 = 66.7% |
6.7 points apart, with equal skill. The weighting follows Overall Equipment Effectiveness (R. C. Hansen, 2001): weight each product by its share of scheduled time.
So compare shifts on the same product, or show the OEE per product per shift. The loss of a hard product belongs to the product and its setup, not the crew.
Step 4: look for real differences
When measurement and mix are fair, real causes remain. First split each shift's stops into planned and unplanned (see planned vs unplanned downtime).
- Support. Maintenance, quality, planners and material handlers are fewer at night and on weekends.
- Staffing. The number of operators, and how many are trained on this line.
- Method. Different crews do the same changeover in different ways. See how to measure changeover time.
- Handover. A shift that starts with a half-prepared line starts with a loss. See shift handover checklist.
- Logging. One crew logs every stop, the other half of them. See operator downtime logging.
Turn the gap into a gain
The best shift is a free teacher. Find out what it does differently, then make it the standard for all.
Overall Equipment Effectiveness (R. C. Hansen, 2001) gives a case where a pilot crew cut its changeover time by 40% and the result spread only slowly to the other crews. Without visible support from the top, a good method spreads slowly.
The same book suggests building an escalation plan with input from the best-performing crews. Their rules become the rules of the plant.
To keep it fair, report OEE by shift next to the loss split every week, and use the numbers for help, not for ranking. See how to calculate OEE for the base maths, and how to reduce unplanned downtime for the actions.
Common mistakes
- Comparing single days.
- Reading OEE only, without the loss split.
- Ranking crews by OEE when they ran different products.
- Different ideal rates or scheduled-time rules by shift.
- Using the gap to punish. People then hide stops, and the next report is worse.
Shift gaps are easy to argue about and hard to see without data. Fabrico is a light MES for OEE: it shows OEE per line and shift, with stops, speed and counts taken from the line and the cause tagged by the operator. See how Fabrico compares OEE per line and shift.
What people ask next
Is it normal for OEE to differ between shifts?
Yes, some gap is normal, because product mix, staffing and support change over the day. A large or growing gap is worth a look. Read the loss breakdown to see where it comes from.
How do I know if the gap is real or a data problem?
Check the ideal rates, the scheduled-time rule, the stop threshold and the code list for each shift. Then compare logged stop minutes with an hour meter. See operator downtime logging.
Should I compare shifts at all?
Yes, to learn, not to rank. Compare the loss split on the same product. Use the best shift's methods as the standard.
Why does the night shift often show lower availability?
Often because support and material services are thinner at night, which shows up as waiting. Check the stop causes with the codes before you assume the crew is the reason.
Can I average the shift OEE values for a daily figure?
Not by a plain average. Weight each shift by its scheduled time, the way Overall Equipment Effectiveness (R. C. Hansen, 2001) does. See the OEE calculator for a quick check.
Related guides: OEE on 12-Hour Shifts: Comparing Crews Fairly · Causes of Reduced Speed on a Production Line