Line cleaning between a gluten batch and a gluten-free batch — the invisible residue the quality manager signs off blind.

The problem

The cross-contact nobody sees until the batch has already been made.

In a food plant that shares a line between gluten product and gluten-free product, the risk is not in the recipe — it is in what stays on the line when the batch changes over:

1. Flour stuck in a crack of the belt — invisible from the walkway, perfectly capable of crumbling onto the first gluten-free product that passes over it.
2. Flour dust settled on the inner wall of a hopper — the operator opens up, looks with a flashlight, sees nothing odd, signs. The camera would have seen it.
3. A bolt gasket with trapped crumbs — the classic point the daily cleaning skims over because "flour doesn't get in there". Until it does.
4. A doser mouth or kneader outlet with a thin crust — cleaned with a cloth, residue left on the unlit edge.

The classic system covers this with three legs: a formal cleaning protocol, ELISA/lateral flow on the first gluten-free batch, and the quality manager's line-release signature. The first two work — but they arrive late, because by the time the ELISA comes back positive, the gluten-free batch has already been made. And the third is where the hole is: one person signing that they have visually verified ten critical points on a 30-meter line, at shift change, without having been able to look inside every gasket.

A single contamination that reaches the shelf is an AOECS recall, a regional fine, a RASFF alert and a gluten-free customer who never comes back. And a "gluten-free" brand is earned over years and lost in a week.

How it fits the IRIS system

iLEAN Edge sees the line after cleaning — and does not let it start until the person signs.

The cross-contact problem on shared lines is not solved by adding another checklist to the quality manager: it is solved by giving them eyes where today they only have the word of the operator who cleaned. iLEAN Edge is exactly that — machine vision cameras that look at every critical point after cleaning, compare it against that line's "clean reference state" and leave the release signature to the person, with point-by-point visual evidence.

Edge looks at the hopper, the belt, the doser, the gasket and the forming table after cleaning. If it sees residue, the line does not start. The person validates and signs — never the other way around.

The iLEAN pieces applied to validating cleaning between a gluten batch and a gluten-free batch:

• Edge — a machine vision terminal (CNN) at every critical point the quality manager defines: mixing hopper, conveyor belt, doser, bolted gasket, forming table, kneader outlet. Each camera learns the "clean reference state" of that point on that line. After cleaning, it fires an inspection and compares — if it detects flour residue, crumbs or settled dust, it blocks the startup and records a photo of the finding. It works without a network: if the plant loses WiFi, Edge keeps inspecting and blocking, because what is critical cannot depend on connectivity.
• Connect — captures the end-of-gluten-batch event (from the ERP, the MES, a manual instruction from the operator or a line marker) and triggers the post-cleaning validation walkthrough. It also captures the quality manager's release signature through whichever channel they use (mobile app, earpiece, tablet), and archives it with the photo and timestamp.
• Agent — orchestrates the full walkthrough: gluten batch end marked → reminds the cleaning crew of the protocol → when they finish, launches the Edge inspection round → if there is residue, holds the line and alerts the quality manager through their channel → when the person signs with the evidence in front of them, releases the startup of the gluten-free batch. And it leaves the per-changeover dossier ready for an IFS/BRC/FSSC 22000 audit.

See the full IRIS architecture →

Before and after

Blind release signature vs. release with point-by-point visual evidence

Aspect	Classic cleaning validation	With iLEAN Edge on the line
Verification after cleaning	Eyeball inspection by the operator and the quality manager	AI vision inspection at every critical point, with an archived photo
Line-release signature	Blind — signed without having looked inside every gasket	With point-by-point visual evidence in front of the manager
Cross-contact detection	Post-production (ELISA on the first gluten-free batch)	Pre-startup (residue seen before the first package)
Cleaning validation time	Varies by shift and workload — sometimes rushed under pressure	Standard walkthrough — Edge does not tire at the end of the shift
Hard-to-reach zones	The assumption that "flour doesn't get in there"	Explicit coverage of the zones the camera can see, reinforced manual protocol where it cannot
File for the IFS/BRC/FSSC auditor	Signed cleaning sheets, reconstructed by hand	Per-changeover dossier: photo + timestamp per point + the manager's signature

Impact estimate

Impact estimate for your plant — to be validated with your numbers.

The block below is an estimate to be validated with the specific data of your plant. We put it forward so the committee has an order of magnitude; we refine it during the diagnostic.

• Food plant with one line shared between gluten product and gluten-free product, 1-2 batch changeovers a day, 6-10 critical points defined by the cleaning plan.
• Edge pilot on the priority critical points (cameras + startup-blocking actuator + integration with the release signature). First value expected within a few weeks: the first post-cleaning inspection with archived visual evidence.
• Indicative payback between 4 and 9 months, depending on the documented average cost of a recall/rework in your sector and the frequency of cross-contact incidents reported in recent years.
• Reduction of cross-contact incidents ≥ 30% against the baseline (the "before" we measure during the immersion) — a defensible floor, not a shiny ceiling.
• The hard lever is a single avoided AOECS recall: product pulled from the shelf, reverse transport, destruction, a potential fine and damage to the "gluten-free" brand. One single recall pays for the pilot many times over.

And the quality manager's reasonable doubt

"What if the AI camera gets it wrong and lets residue through, or worse, says there is residue when there isn't and stops the line for nothing?" — hallucination is a problem of free generation, not of anchored tasks. In tasks where the AI merely compares the current image with a previously validated "clean reference state" (which is exactly what Edge does here), the best models brought error below 1.5% ^[1]. And even so, what is critical is never decided alone: Edge holds the line and the person signs, looking at the photo that triggered the hold. iLEAN's three safety rings are there precisely for this — the AI proposes, the quality manager decides.

^[1] OpenAI paper "Why Language Models Hallucinate", 2025 — on the reliability of AI in anchored tasks.