OEE for High-Mix Food Lines: Setting a Fair Target

Key takeaways

• The famous 85% target comes from Nakajima's Introduction to TPM (1984), built on high-volume lines with few changeovers.
• Most plants score closer to 60%. A high-mix food line is not failing because it sits below 85%.
• Changeovers cap Availability on a high-mix line, and many of them include validated cleans.
• Set the target per line, from its own baseline and the losses it can control.
• Report OEE by SKU, so the formats that cause the most stops stand out.

A food line that switches products three times a day, with allergen cleans between them, has a lower ceiling. It should not chase the OEE of a line that runs one product all week. Its fair target comes from its own changeover load, its own baseline and the losses it can actually control.

Fabrico, OEE and light MES software with AI cameras, measures OEE live by line and shift and prices each loss in money. That makes it possible to compare lines on what they lose, not only on one percentage.

Why is 85% the wrong target for a high-mix food line?

The 85% figure comes from Seiichi Nakajima's Introduction to TPM, published in 1984. Its roots are Japan in the 1970s, in automotive and discrete production. It is usually split as 90% Availability, 95% Performance and 99% Quality.

That benchmark was built on high-volume, dedicated lines running one product with few changeovers. It misleads on high-mix lines with allergen changeovers. Most manufacturers score closer to 60%. One study of more than 3,500 machines in over 50 countries found OEE peaks at 55 to 60%.

So a well-run high-mix food line can sit well below 85% and still be doing a good job. Our overview of OEE benchmarks by industry shows how much the numbers vary.

How much do changeovers cap a high-mix line?

Every changeover sits inside planned production time, so it lowers Availability. On a high-mix line, the changeovers alone set a ceiling the line can never pass, however well it runs between them. Part of each changeover is often a validated clean, which cannot be shortened without revalidation.

Worked example: two lines that run equally well

Sample line (illustrative numbers, not a customer). Two lines each have 960 planned minutes a day. Both lose 40 minutes a day to breakdowns, both run at 92% Performance and 98% Quality. Line A is dedicated to one product. Line B is high-mix, with three changeovers a day of 60 minutes each.

The two lines are run equally well. The whole gap of 16.9 points is changeover time. Even with zero breakdowns, Line B could not pass 780 / 960 = 81.3% Availability.

So where is Line B's real loss? Its changeovers average 72 minutes against a 60 minute standard. That is 3 x 12 = 36 minutes over standard each day.

At 200 units a minute and a margin of USD 0.20 a unit, a lost minute costs USD 40. Over 250 days, the overruns cost 36 x 250 x USD 40 = USD 360,000 a year (about EUR 331,000).

That number is Line B's fair target. It does not touch a validated clean.

How should you set a fair OEE target?

1. Measure a signed baseline for the line before you set any target.
2. Split the losses into what validation fixes, such as clean cycles, and what the team controls: overruns, waiting, breakdowns, speed and rejects.
3. Target the controllable part, such as halving the minutes over the changeover standard.
4. Compare lines on losses in minutes and money, not on OEE alone.
5. Review the target when the product mix changes, because the ceiling moves with it.

Our guide to setting OEE targets by line goes deeper on the method.

Why report OEE by SKU on a high-mix line?

A line OEE averaged over all products hides the two or three formats that cause most of the stops. Report OEE and stop minutes by SKU, using the right ideal speed for each one. Then the team can see which product switch, or which pack size, costs the most.

Speed needs care too. Nameplate speed can be unfair for some SKUs, while letting a line run slowly for months hides a real loss. Agree the ideal speed per SKU once, and write it down. For more on reducing the minutes, read our guide to changeover time.

Which mistakes do high-mix plants make with OEE?

• Comparing a high-mix line with the 85% benchmark from dedicated lines.
• Hiding changeovers in planned time with no standard, so overruns never show.
• Reporting OEE only by line, never by SKU.
• Shaving a validated allergen clean to win OEE. That is a food safety breach.
• Letting changeover gains slide back because the new standard work was never written down.

Where does Fabrico fit?

Fabrico is a computer vision enabled Manufacturing Performance Platform. OEE and light MES come first, on one time model for every line.

• Automatic downtime logging, so each changeover is timed from the machine signal.
• The operator confirms the reason in one tap, so allergen cleans and format changes get their own codes.
• Losses ranked by money, so changeover overruns are priced next to breakdowns.
• OEE live by line and shift, and one view across plants.

It is the OEE layer of a light MES. You start on one line, with a baseline and a decision date agreed in advance. There is no free plan and there are no tiers: each plant gets a quote. For more on high-mix production, see our high-mix, low-volume guide.

Frequently asked questions

What is a good OEE for a food production line?

There is no single number. Most manufacturers score closer to 60%, and one large study found OEE peaks at 55 to 60%. A fair target for your line comes from its own baseline, its product mix and the losses it can control, not from a general benchmark.

Is 85% OEE realistic for a high-mix line?

Rarely. The 85% figure comes from Nakajima's work in 1984, based on high-volume lines with few changeovers. A line that switches products several times a day, with cleans between them, loses a large share of its time to changeovers it cannot avoid.

How do changeovers affect OEE?

Changeovers sit inside planned production time, so they lower Availability. On a high-mix line they set a ceiling: with three changeovers of 60 minutes in 960 planned minutes, Availability cannot pass 81.3%. Track the minutes over each changeover's standard to find the controllable loss.

Should OEE targets be set per line or per plant?

Per line. Lines with different product mixes, changeover loads and speeds cannot share one fair target. Set each line's target from its own baseline. At plant level, compare lines on losses in minutes and money, and sum minutes rather than averaging percentages.

How do you compare a high-mix line with a dedicated line?

Compare what each one loses, not only their OEE. Split each line's losses into changeovers, breakdowns, speed and rejects, in minutes and money. Two lines that run equally well can still differ by many OEE points because of changeover load alone.

Find your line's fair target

Bring one month of changeover and stop data. We will show your line's ceiling and the minutes you can win.

Request a demo · Contact us