The lot travels on its own from order to bag

The coder that prints lot and date on every 50 lb bag and every big bag is relatively modern and has an API. So does the weigh-bagger. But both are isolated from the ERP: at every changeover someone types the new lot and reference into the panel, by hand, on the floor, wearing gloves. With iLEAN they are stitched to the ERP and lot, date, product, target weight and special marking travel automatically from the active order.

‹ See all cases of cane sugar refining

Coder and weigh-bagger on a sugar bagging line marking lot and date on 50 lb bags, with the operator confirming the active order on a panel
The problem

The coder is the last physical piece before the customer. And it is still typed by hand.

The coder is the last physical piece before the customer. Any error in lot, date or marking there becomes a downstream traceability problem. And yet the standard method is still typing on the panel, because "integrating it properly takes time and means stopping the line". In a plant designed with few machines and many different qualities, that typing repeats many times a day. The probability of error is not a hypothesis: it is a function of frequency. One wrong digit and half a production run goes out with the wrong lot. In a product shipped in bulk to bottling and baking plants, a wrong lot is not an ugly label: it is broken traceability that forces you to widen the scope of any recall far beyond what was necessary.

  • Any error in lot, date or marking there becomes a downstream traceability problem.
  • And yet the standard method is still typing on the panel, because "integrating it properly takes time and means stopping the line".
  • In a plant designed with few machines and many different qualities, that typing repeats many times a day.
  • The probability of error is not a hypothesis: it is arithmetic.
How it fits the IRIS system

Connect stitches two modern systems via API, replacing neither.

Connect stitches two modern systems via API, replacing neither:

The product changeover goes from a sequence of keystrokes on two separate panels to one tap in the ERP. And no new middleware is introduced: the interfaces the machines already ship with are the ones used.

  • When the ERP flags the order start, it automatically publishes product, lot, date and special marking to the coder.
  • In parallel it publishes target weight and tolerances to the weigh-bagger.
  • The product changeover goes from a sequence of keystrokes on two separate panels to one tap in the ERP.
  • No new middleware to maintain is introduced: the interfaces the machines already ship with, and that sit unused today, are the ones being used.

See the full IRIS architecture →

Before and after

The coder, before and after

AspectTodayWith iLEAN Connect
Lot, date and markingTyped on the panelPublished from the active order
Target weight and tolerancesTyped on the baggerPublished in parallel
A product changeoverKeystrokes on two panelsOne tap in the ERP
Risk of wrong markingArithmetic, with many qualities a dayRemoved at source
New middleware to maintainNone
Line stoppage to integrateThe reason it never gets doneNot required

Impact estimate

Impact estimate — to validate against 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.
  • Drastic reduction in the risk of broken traceability from wrong marking.
  • It is also the technical prerequisite for ring 2 of the complete system: without the lot travelling on its own, the rings cannot be coordinated.
  • And a changeover that stops depending on who is on the panel.

estimated payback of 3 to 6 months, with a drastic reduction in the risk of broken traceability from wrong marking. It is also the technical prerequisite for ring 2 of the complete system (case 12). *Estimate to be validated* after confirming the installed models expose an API.

And the fair question from the production manager

"Does this mean another system to maintain?" — no, and that is deliberate. No middleware is introduced: what gets used are the interfaces the machines already ship with and that are already paid for. The work is stitching them to the active order, not adding a layer between them.

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

Frequently asked questions

What people ask about integrating the coder

Does the line have to stop to integrate it?

No, and that is precisely why it never got done before: the perception that it means a shutdown. The interfaces are already there; what is missing is somebody publishing the active order into them.

Does it work with coders and baggers from different makers?

Yes, and it is the usual case. Each machine is stitched its own way, but what is published to them comes from the same place, which is the active order.

What happens if the ERP order is wrong?

Then the wrong marking gets published, and it is better to find that out here than at the customer. In practice this surfaces order-master errors nobody knew existed.

Who confirms the changeover?

The line operator, with one tap instead of two panels' worth of keystrokes. The confirmation stays human; what disappears is the typing.

Why is this a prerequisite for the full system?

Because ring 2 coordinates identity with process, and identity has to travel with the bag on its own. If the lot is typed, the chain has a manual link and the whole ring rests on it.

Let's talk

Tell us how many times a day somebody types a lot on your coder panel.

We work on your plant's real data, not ours. Assessment with no commitment.

Request estimated ROI within 48h ‹ See all cases of cane sugar refining See food