The delivery note that doesn't wait

The problem

Traceability starts with someone typing while a truck waits.

While the driver waits, racking up demurrage, someone would have to manually enter every line of the delivery note into the ERP/WMS. In a multi-client warehouse that can end up being shipped to the wrong client weeks later, with nobody catching it in time.

• Every inbound truck brings a paper delivery note and lot and expiry labels on every pallet: the data that makes traceability possible for food and medical clients.
• While the driver waits, racking up demurrage, someone has to enter every line of the delivery note into the ERP/WMS by hand, with the next truck already queuing at the gate.
• A lot or expiry typed wrong at receiving does not fail at receiving. In a multi-client warehouse it can ship to the wrong client weeks later, or break first-expired-first-out picking.
• And the receiving module needs 2-3 active ERP licenses just so those lines can be typed, whether the warehouse is busy or not.

How it fits the IRIS system

Connect in photo mode at receiving — delivery note and labels, no supplier changes.

The vision AI recognizes the delivery note's layout without asking the supplier to change theirs, and reads GS1-128, QR, barcode or plain OCR on the labels. The receiving manager confirms on the industrial tablet; the data crosses over automatically via API.

Suppliers will not change their delivery note for you, and they do not have to. The vision model recognizes each supplier's layout as it comes and reads the labels with whichever code they carry. The receiving lead stops typing and starts checking, and the truck leaves the dock with its data already in the ERP.

See the full IRIS architecture →

Before and after

Today's receiving versus photographed receiving

Aspect	Today	With iLEAN Connect
Delivery note entry	Typed line by line	One photo
Lot and expiry	Typed from the label	Read from GS1-128, QR, barcode or text
Driver waiting at the dock	Demurrage running	Data in the ERP within minutes
Supplier's document format	Whatever they send	Recognized as it is
Wrong lot at receiving	Discovered weeks later, at shipping	Stopped at the validation tap
Receiving module licenses in the warehouse	2-3 active	Only whoever validates

Before: the truck waits racking up demurrage while data is typed line by line, risking lot errors. After: photo, one-tap validation, data in the ERP within minutes.

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 3-8 months, depending on inbound volume and the number of receiving licenses today.
• -50 to -70% on receiving/warehouse module licenses: only the receiving lead validates with a tap.
• Less demurrage, because the truck no longer waits for line-by-line data entry before it can leave the dock.
• And lot and expiry enter correct from the start, which is what protects the shipment weeks later and keeps expiry-based picking honest.

Payback 3-8 months · -50 to -70% on receiving/warehouse module licenses. The receiving module today requires 2-3 active licenses; with Connect, only the manager validates with a tap — the rest no longer need direct ERP access.

And the fair question from the production manager

“What if it misreads a lot number on a creased label?” — reading a delivery note and a coded label is an anchored task, where the best models drop below 1.5% error [1], and a barcode or GS1-128 read is checked by its own structure. When a label is only legible as text and confidence drops, the field is highlighted on the tablet for the receiving lead to confirm against the pallet in front of him. Every confirmed line keeps the photo it was read from, so a lot question weeks later is answered by looking at the label itself.

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