Sanitation-Adjusted OEE for Indian Food Lines

On a food line, a chunk of every shift is not production — it is cleaning: wash-downs, CIP (clean-in-place), allergen change-cleans, pre-start sanitising. If you leave that time inside your OEE denominator, your number looks terrible and tells you nothing. If you quietly dump it all into "planned downtime," you hide the real problem — the day the 60-minute CIP took 95. The honest answer is in between: exclude necessary sanitation from available time, but measure the overrun as its own loss. That is sanitation-adjusted OEE.

By the Mama Editorial Team. Standard OEE was built for discrete machines that run until they break. Food lines stop on purpose, every shift, for hygiene that the law and your buyers require. This article shows how to count that time so OEE stays honest — and where the hidden rupees are.

Key takeaways

• Raw OEE misleads on a food line because mandatory sanitation eats scheduled time that was never available for production.
• Treat necessary sanitation as planned non-production — exclude it from the OEE Availability denominator, the same way you exclude breaks and no-demand time.
• But track the overrun separately. A CIP that runs 95 minutes against a 60-minute standard is a 35-minute loss that vanishes if you bury it in "planned downtime."
• Report two numbers: OEE on the true production window, and a sanitation schedule-adherence metric (standard time ÷ actual time).
• Sanitation is required, not optional — FSSAI Schedule 4 hygiene and your buyer audits. The goal is not less cleaning; it is cleaning that starts on time and finishes on time.
• A camera times the wash honestly; it cannot certify the clean. Timing is OEE's job; the swab/ATP result stays with QA.

Why raw OEE breaks on a food line

Recall the shape of OEE. You start from a block of time, subtract what was never meant to make product to get Planned Production Time, then OEE = Availability × Performance × Quality measured inside that window.

On a discrete machine the only things you subtract are breaks and genuine no-demand. On a food line you must also subtract mandatory sanitation — the pre-start clean, mid-shift wash-downs, end-of-run CIP, allergen change-cleans. That time was never available to make product; the line is supposed to be stopped and wet.

Two wrong ways to handle it:

• Leave sanitation inside the denominator. Now a line that cleaned exactly as required scores a low "Availability" as if cleaning were a breakdown. The number punishes you for doing the mandatory thing, and managers learn to ignore it.
• Dump all sanitation into "planned downtime" and forget it. Now the denominator looks right, but you have made the single most controllable food-line loss — cleaning that overruns — completely invisible. A CIP that balloons from 60 to 95 minutes looks identical to one that finished on time.

The fix: necessary vs overrun

Split sanitation into two buckets:

Then report two numbers, not one:

1. Sanitation-adjusted OEE — Availability × Performance × Quality measured on the true production window (after necessary sanitation is removed). This tells you how well the line ran when it was meant to be running.
2. Sanitation schedule-adherence — standard sanitation time ÷ actual sanitation time. 60/95 = 63%. This tells you how well the cleaning was run. Keep it beside OEE, never inside it.

This mirrors the discipline in our note on reading OEE benchmarks honestly: one combined number hides the lever; two honest numbers point at it.

Worked example, both ways, in rupees

One 8-hour shift. Illustrative numbers — substitute your own.

Now the shift as it actually ran:

• CIP overran: 95 min instead of 60 → 35 min overrun
• One unplanned re-clean after a product spill: 20 min
• Inside the remaining run window, normal Availability/Performance/Quality losses

The two numbers:

• Sanitation schedule-adherence = 60 ÷ (95 + 20) = 52%. Cleaning took nearly twice its standard time. This is the headline problem — and a raw OEE would never show it.
• Sanitation-adjusted OEE: the 35-min overrun + 20-min re-clean (55 min) is stolen from the 390-min planned window, so real run time starts 55 min short before any machine loss. Measure Availability × Performance × Quality on what is left and you get a figure you can trust — and a clear culprit.

The rupees. Value the line at its contribution margin per running minute — say 25 packs/min × ₹2/pack = ₹50/min:

At two shifts a day, 300 days: ~₹16.5 lakh a year — from sanitation overrun alone, on one line, entirely hidden by a raw OEE that called it all "planned downtime."

Sanitation is required — the point is to run it tight

None of this is an argument to clean less. Hygiene on a food line is mandatory: FSSAI's Schedule 4 (the GMP/GHP annexe to the Licensing & Registration Regulations, 2011) requires cleaning and sanitation of equipment and premises, and your retail and export buyer audits check it directly. Dairy and wet-process lines live on CIP — see cameras in Indian dairy plants and the broader food-plant hygiene and audit picture.

So the target is not fewer cleans. It is cleaning that starts on time, follows the standard, and finishes on time — and a line that is ready to run the moment the clean is signed off, not one sitting wet while the crew drifts back. Sanitation-adjusted OEE makes that discipline visible without ever pressuring anyone to cut a clean short.

What a camera can — and cannot — do

Be precise here, because food hygiene is exactly where overselling is dangerous. A camera is good at timing: it sees when the line stopped for cleaning, when the crew actually started, when the wash ended, and whether the line then sat idle before production resumed. That is precisely the data sanitation-adjusted OEE needs — the plan vs actual of the clean, which no shift book captures reliably.

What a camera cannot do: certify that the clean was effective. It does not read a swab, an ATP reading or an allergen test. Pack weight and seal integrity are likewise not its job. A camera times the wash; it cannot certify the clean. The hygiene result stays with QA and the lab.

That is the honest bridge for Mama: from your cameras (or ones we place) it watches the floor and every morning sends the owner, on WhatsApp, where the hours went — including sanitation that started late or ran long, and the line sitting idle around it — with a rupee estimate. It will not tell you whether a surface passed a swab; it will tell you the ₹ sitting in your cleaning overruns. Send a short video of the line and we will return where the hours go, a camera plan, and a proposal.

How a plant head should start

• Set a standard time for each clean and CIP. Without a standard there is no overrun to measure.
• Exclude necessary sanitation from the OEE denominator — and write the convention down so nobody thinks the number is wrong.
• Track schedule-adherence (standard ÷ actual) beside OEE, never inside it.
• Put the overrun in rupees at margin/min so the floor sees the cost of a wet line waiting.
• Keep the clean sacred — the goal is a tighter schedule, never a shorter wash.

FAQ

How do you calculate OEE on a food line that has daily cleaning? Subtract necessary sanitation (at its standard duration), along with breaks and no-demand, to get Planned Production Time, then compute Availability × Performance × Quality inside that window. Track any sanitation overrun separately as a loss — don't leave it in the denominator and don't bury it in planned downtime.

Should CIP time count against OEE? Standard, required CIP should be excluded from the OEE Availability denominator, like any planned non-production. But the overrun — minutes beyond the standard, plus re-cleans and idle-wet waiting — should be measured as a named loss, because that part is controllable.

What is sanitation schedule-adherence? Standard sanitation time ÷ actual sanitation time. A 60-minute CIP that runs 95 minutes scores 63%. Report it alongside OEE so cleaning overruns stay visible.

Does measuring this mean cleaning less? No. FSSAI Schedule 4 and buyer audits require the hygiene. The goal is cleaning that starts on time and finishes on time, and a line ready to run the moment it is signed off — not a shorter or skipped clean.

Can a camera measure sanitation time? Yes — it times when cleaning started, ended, and whether the line sat idle afterward. It cannot certify the clean was effective; swab, ATP and allergen results stay with QA.