The first hour often loses output because the line starts late after handover, restarts cold, waits for material or sign-offs, and runs slowly until it settles. To see it, line up hourly good counts by hour of shift over at least 20 shifts and compare hour 1 with your steady hours. Then split the gap into late start, stops, slow running and scrap.
Ask a supervisor when the shift starts and you will hear a clock time. Ask the line, and the counter may say something quite different.
An eight-hour shift that loses a third of its first hour every day gives away about 20 minutes, and it does so quietly, because the shift still has seven more hours to look fine in.
How to see the pattern
A daily OEE number cannot show this. You need counts per hour, and you need them lined up by hour of shift rather than by date.
One line, one product family if you can, for at least 20 shifts.
Hour 1 is the first hour after the bell, whatever the clock says.
A median stops one bad breakdown from bending the shape.
Leave out hours with a planned break.
If you do not have counts per hour yet, start an hour-by-hour board. A few weeks of it will draw the same shape.
Example: a line with an ideal rate of 120 units a minute, so 7,200 an hour. Hour 5 includes a 15-minute planned break.
The steady hours make 5,760 units. Hour 1 falls 2,160 short of that, which is worth 2,160 / 120 = 18 minutes at the ideal rate.
Hour 2 adds another 360 units, or 3 minutes. Together that is 21 minutes a shift, or 4.4% of 480, on top of whatever the line loses in any normal hour.
Hour 8 has a dip as well, which has its own causes. See end of shift losses.
Split the first hour into its parts
A low count alone will not tell you what to do. Now take the first hour apart, using the time of the first unit, the stop list, the total count and the scrap.
Example: the first unit arrives 10 minutes after the bell, the line then stops for 6 minutes, and it makes 3,840 units in the remaining 44 minutes. Of those, 240 are scrapped.
Compare that with a steady hour here, which loses 12 minutes (60 - 5,760 / 120). So 18 of the first hour's 30 lost minutes are special to the shift start.
Two simple numbers are worth tracking from now on: minutes to first good unit, and minutes to steady rate. Both are easy to read and hard to argue with.
Why the first hour loses output
| Cause | Sign in the data | What helps |
|---|---|---|
| Line stopped for handover | No units around the bell, on every crew | Hand over at the running line, with an overlap |
| Meeting held off the floor | Fixed late start, same length each day | Short stand-up at the line, or stagger it |
| Cold restart | Slow rate and scrap, worse after a weekend | Warm up before the shift starts, with a set routine |
| Material not staged | Waiting stops in the first 30 minutes | Last shift stages the next job |
| Crew still being assigned | Waiting stops, or slow running with few people | Post the plan before the bell |
| Sign-offs and first-piece checks | Stop until someone arrives | Agree who checks, and by when |
| Problem left by last shift | A fault or changeover carries over the bell | A handover that names open problems |
Causes often chain together, so a late start hides a waiting stop that hides a cold machine. Work on the biggest slice first, then look again.
Compare before you conclude
- Crew against crew. If all crews show the same shape, it is the system, not the people. Why OEE differs between shifts has the checks.
- Monday against midweek. A much worse Monday points to cold starts, which startup losses covers in detail.
- Running through against restarting. Does the line keep running across the bell? If it does, and the dip is still there, the cause is not the restart.
- Product. A planned changeover at the start of shift can look like a late start. Tag it as a changeover.
A note on the rules
If your plant takes a start-of-shift meeting out of scheduled time, this loss partly disappears from OEE. It does not disappear from your output.
Either rule can be applied honestly, as long as it is written down and used for all crews. See how to count breaks in OEE.
How to shrink it
Start with the cheap, boring fixes. Most need agreement between crews, and no money at all.
- Agree a start target. For example, first good unit within 5 minutes of the bell. Put it on the board.
- Hand over at the line. The shift handover checklist keeps it short and covers open problems.
- Prepare in the last shift. Material staged, tools back, the next job set. That links directly to waiting for material.
- Check the first hour first. A quick look at the first 30 minutes' count catches trouble early, and that is the idea behind short interval control.
- Discuss it once a day. Put minutes to first good unit on the agenda of the daily production meeting.
Be careful with one fix. Overlapping crews costs paid time, so compare that cost with the output you win.
What is it worth?
Example: suppose the line above wins back half of the 21 minutes on each of 3 shifts. That is 31.5 minutes a day, or 3,780 units at the ideal rate.
Put your own margin into the hidden factory calculator to see the yearly figure, then test the change for two weeks and compare hour 1 before and after.
This pattern is easy to see once counts are in place. Fabrico is a light MES for OEE that takes stops, speed and counts live from your machines, shows OEE per line and shift, and lets operators tag the cause at the line. See how Fabrico shows OEE per shift.
What people ask next
How much output is normally lost in the first hour?
We found no neutral benchmark for this, and the size depends on the process and how crews change over. Measure your hour 1 against your own steady hours. That gap is the number to work on.
Does the first hour count in OEE?
Yes, if it is scheduled time, it counts like any other. Late starts lower availability, slow running lowers performance and scrap lowers quality. Check your rule for meetings in how to count breaks in OEE.
Is a late start downtime?
If the line was scheduled to run and did not, yes. Log it under its own reason, such as "shift start", so it stops being a mystery. Downtime reason codes helps you set that up.
Why is Monday morning worse than other days?
A line that has been idle for a weekend usually restarts cold, may need cleaning checks, and may be waiting for material. Compare Monday's first hour with Tuesday's to size the effect. Startup losses shows how to count it.
Can we fix this without new software?
Yes. An hour-by-hour board, a written first-unit target and a short handover will get you a long way. Software helps when you want the same view across many lines without manual work.
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: OEE on 12-Hour Shifts: Comparing Crews Fairly · OEE on a New Line: When to Start the Clock · Shop Floor KPIs: What the Crew Should See · Manufacturing KPI Board: Layout That Gets Used