OEE for a press is availability x performance x quality: makeready and stops reduce availability, running below rated speed reduces performance, and spoiled sheets reduce quality. Count in press units, such as sheets, impressions or metres, and keep a fixed speed reference, so that the number stays comparable across jobs.
Why a press needs its own rules
A press changes jobs more often than most production equipment. Short runs mean the machine can spend a large share of the day getting ready, with waste paper to show for it.
Speed also moves with the stock, ink coverage and the operator's comfort. So three questions decide whether your press OEE means anything: where makeready goes, how waste is counted and what "full speed" is.
The core definitions are in what is OEE, and this page applies them to print.
Where each press loss goes
| Press event | OEE factor | Note |
|---|---|---|
| Makeready: plates, ink, register, color to first OK sheet | Availability | Planned, but still lost press time |
| Blanket or plate wash during a job | Availability | A stop, whatever its reason |
| Web break, feeder fault, breakdown | Availability | Over your stop threshold |
| Short sheet jams and double sheet stops | Performance | Under your stop threshold |
| Running below rated speed | Performance | Stock, coverage or habit |
| Sheets spoiled after the first OK sheet | Quality | Color drift, marks, hickeys |
| Sheets used during makeready | Not an extra OEE loss | Time already counted; track as waste |
| No job scheduled | Outside OEE | See TEEP for this |
Makeready: count it, and do not exclude it
Some shops leave makeready out, because it is planned and the customer pays for it. That can hide the biggest lever on a short-run press.
In OEE, a changeover inside scheduled time is an availability loss. See how to count changeovers in OEE for the reasoning.
Define its end clearly, or the numbers will be argued over. A common rule: makeready runs from the last good sheet of the old job to the first sellable sheet of the new one.
Worked example: one shift on a sheetfed press
Example: a hypothetical offset press has a rated speed of 15,000 sheets an hour, or 250 a minute. It is scheduled for 480 minutes and runs 4 jobs.
| Measure | Value | Sum |
|---|---|---|
| Scheduled time | 480 min | |
| Makeready | 100 min | 4 jobs x 25 min |
| Wash-ups and other stops | 38 min | |
| Run time | 342 min | 480 - 100 - 38 |
| Sheets run after OK | 66,000 | |
| Sheets spoiled after OK | 900 | |
| Good sheets | 65,100 | 66,000 - 900 |
| Availability | 71.3% | 342 / 480 |
| Performance | 77.2% | 66,000 / (342 x 250) |
| Quality | 98.6% | 65,100 / 66,000 |
| OEE | 54.3% | Check: (65,100 / 250) / 480 |
Actual speed here was 66,000 / 342 = 193 sheets a minute, against 250 rated. That is about 11,600 sheets an hour on a 15,000 press.
Enter your own figures in the OEE calculator to compare.
Waste: keep two separate numbers
In this example, each makeready also used about 300 sheets, 1,200 in total. They cost money, and they must not vanish from your records.
They do not reduce OEE a second time, though, because the time they took is already in the 100 minutes. Report them as a waste measure beside OEE.
- Makeready waste: 1,200 sheets, or 300 a job.
- Running waste: 900 sheets, or 1.4% of sheets run.
- Total waste rate: (1,200 + 900) / (66,000 + 1,200) = 3.1% of all sheets fed.
Some plants count all sheets in quality and count run-up time as running. That can also be valid, but pick one rule, as mixing them counts the same loss twice.
Use how to calculate scrap rate for the waste side.
Speed: choose a fixed reference
The ideal rate is the biggest choice on a press. The stock and job may slow the machine, but if your reference drops with each difficult job, the speed loss disappears from view.
- Rated speed of the press: simple and hard to game. The score will vary with the job mix, so read the trends by stock type.
- A fixed standard per stock class: fairer between jobs, with less bias from job mix. Set it from demonstrated best speed, and do not lower it to match a bad week.
Whichever you choose, never set the reference from the average. An average makes performance look close to 100%, as ideal cycle time explains.
Use press units too: cut-offs or metres on a web press, metres on a flexo line, and sheets or impressions on a sheetfed press. Finished copies per sheet are the wrong unit.
How job length moves the score
Short runs are a fact of life in print, and they change availability a lot. Here is the same press, with the same 25-minute makeready and 38 minutes of other stops.
So do not compare a commercial short-run press with a long-run publication press on OEE alone. Compare each press with itself, or look at makeready per job, as discussed in OEE with mixed products.
The way forward is to cut makeready, because you may not control the job mix. Cutting 25 minutes to 15 on the 8-job day gives back 80 minutes, moving availability from 50.4% to 67.1%.
Prepare what you can before the stop: plates, ink, stock and job data. SMED changeover reduction gives the method, and how to measure changeover time shows how to time the phases.
Getting the data off a press
Many newer presses log job data in the console, and some exchange job and status data with print management systems through CIP4 formats such as JDF and its messaging part, JMF. Ask your press maker what your model and options allow.
An older press can still give you what you need:
- Sheet count: a sensor at the delivery, or the machine's own counter, if it can give a pulse. See part counting sensors.
- Run state: a motor or drive signal, combined with that count. See machine state detection.
- Job boundaries: an operator entry or a scan of the job ticket, to mark the start of makeready and the first OK sheet.
- Waste: a tally of spoiled sheets from the operator, a waste gate counter or a sheet count at the delivery pile.
Keep guards in place when you fit a sensor, and leave any wiring in the press cabinet to a qualified electrician.
Those job marks do more than anything else. Without a clear first-OK mark, makeready and running blur together, and neither number can be trusted.
Mistakes press teams make
- Using copies instead of sheets. A 16-up layout multiplies output, and it tells you nothing about the press.
- Lowering the reference for hard jobs. Performance looks steady, and real speed loss goes unseen.
- Treating wash-ups as nothing. They are press stops, so record how often they happen and how long they last.
- Leaving all waste to a final count. A count taken at the end of a job cannot show you which shift or phase wasted the sheets.
- Chasing one big OEE score. Use the factor that is worst, since each one has a different fix. OEE calculation mistakes lists more.
Your first two weeks
Time every makeready for two weeks, broken into plates, ink and color, and register. Those phases usually show what to fix first, and that is often the fastest way to raise press OEE.
Makeready, wash-ups and breakdowns all look like stops until a person names them. With Fabrico, a light MES for OEE, stops, speed and counts come live 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. See how Fabrico tracks stops.
What people ask next
Is makeready part of OEE?
Yes. It takes place inside scheduled time, so it reduces availability even though it is planned. Keep it visible as its own reason, as in how to count changeovers in OEE.
What is a good OEE for a printing press?
We found no neutral published benchmark specific to printing presses, and job length changes the answer a lot. Compare each press with its own past results, at similar job lengths. How to set OEE targets shows how to build a target from your own data.
Should I use rated speed or actual job speed for performance?
Use a fixed reference, usually rated speed or a demonstrated best speed per stock class. Do not reset it to each job's actual speed, as this hides the loss. See how to calculate OEE performance.
Do makeready sheets count as quality loss?
Under the rules on this page they do not, because the time is already counted in availability. Track them as a separate waste number so that the cost stays visible. If you use another rule, make sure the same loss is not counted twice.
How do I calculate OEE for a web or flexo press?
The method is the same, but you count metres or cut-offs instead of sheets. Performance is actual metres divided by run time x reference speed in metres a minute. Quality is good metres divided by total metres.
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.