The modified-atmosphere packer, stitched to the ERP

The problem

The same data gets typed twice, right before the product ships.

The modified-atmosphere packer is the last physical step before product ships to the customer. It has a modern API, but it's isolated from the ERP because "integrating it properly takes time." Any typing error for batch, date, or allergen on its panel becomes a recall.

• The modified-atmosphere packer is the last physical step before the tray ships to the customer: what it prints there is what the consumer will read and what a retail chain will claim against if it does not add up.
• It has a modern, documented API, but it sits isolated from the ERP because integrating it properly never reaches the top of the priority list: there is always something more urgent than connecting two systems that, more or less, already work separately.
• One wrong finger on that panel is a recall: a tray labeled without declaring dairy, or with an expiry that does not correspond, is not fixed with rework — it is fixed by calling the customer and pulling product off the shelf.
• And the risk grows with the number of references: the more formulas and formats rotate through the same packer — MAP tray, bulk for foodservice, frozen bag for retail — the more times a day somebody hand-types the data that defines what the consumer will read on the pack.

How it fits the IRIS system

Connect stitches two modern systems by API, replacing neither.

Connect stitches two modern systems via API: when the ERP marks the order's start, it automatically publishes batch, date, and allergen list to the packer. A batch change becomes one tap in the ERP, not typing on the machine.

The batch change goes from typing on the machine to one tap in the ERP. And what was printed comes back confirmed to the system: not what was asked to be printed, but what the packer reports having printed on that tray. That distinction is what turns coding into evidence instead of an assumption.

See the full IRIS architecture →

Before and after

The isolated packer versus the stitched packer

Aspect	Today	With iLEAN Connect
Typing batch and expiry	Twice per changeover	Once, in the ERP
Printed allergen list	Depends on the panel	Travels from the active order
Labeling recall risk	Permanent	Eliminated at source
Format or formula change	Minutes on the panel	One tap in the ERP
What was actually printed	Assumed	Confirmed and returned to the system
The two systems	—	Neither is replaced

Recurring typing errors, some triggering recalls → zero human-origin coding errors.

Impact estimate

Impact estimate — to be validated with your numbers.

The block below is an estimate to be validated against your plant's actual data. We put it forward so the committee has an order of magnitude; we refine it during the assessment.

• Estimated payback 3-6 months, counting only the changeover time recovered on the panel across every batch change in the shift.
• A drastic reduction in mislabeling recall risk, which is where the case's real value sits.
• Zero human-origin errors in coding batch, expiry and allergens, on MAP trays as well as bulk cases.
• And with many references rotating on the same line, the number of batch changes stops increasing your exposure to error.

Estimated payback 3-6 months, with a drastic reduction in mislabeling recall risk. *Estimate to validate*.

And the fair question from the production manager

“Isn't a classic integration with the machine builder cleaner?” — it is, the day the project reaches the top of the list, and that is exactly the problem: this is the textbook case of a modern machine still isolated because there was never time. This stitch replaces neither system and asks for no downtime window. Publishing the data is besides an anchored task, with the best models below 1.5% error [1] in reading the order.

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