The steel coil enters the ERP with a photo, not with typed lines
The truck arrives with 6-10 coils. The receiving manager photographs the A3 delivery note and each coil tag, and iLEAN Connect extracts heat, gauge, width and supplier by OCR. After a one-tap validation, everything enters the ERP with coil traceability intact. The truck stops blocking the dock.
Coil receiving is where the dock gets blocked — and where the wrong heat is born that blows up at the press weeks later.
In a Tier-2 automotive metal stamping plant, every truck arrives with 6-10 steel coils, its A3 delivery note and one tag per coil, and today all of that entry goes through the receiving manager's keyboard:
- The steelmaker A3 delivery note — heat, gauge, width and supplier typed row by row into the receiving module while the truck waits 30-45 minutes charging demurrage at the dock.
- Every coil tag — the heat number copied in haste to release the carrier as soon as possible, exactly when getting a digit wrong is easiest.
- Cross-checking against the order — gauges and widths verified by hand against the purchase order, with the carrier waiting on the record to leave.
And if the receiving manager types fast, a wrong heat gets in. The error is not visible at the dock or in the warehouse: that coil, when the stamping line starts three weeks later, shows up with properties different from the ones expected — yield strength, thickness or a mill certificate that do not match the order — and the problem blows up at the press, not at receiving. And the coil-heat-lot-part traceability chain breaks when the delivery note gets misfiled between trucks — precisely the document you have to be able to show if each OEM's requirements lead to an investigation into the origin of a material under IATF 16949.
Connect in photo-over-analog mode — A3 delivery note and coil tags into the ERP with heat traceability, no typing.
Coil receiving does not need a new workstation at the dock or steelmakers changing their paperwork. It needs the A3 delivery note that already arrives in the truck cab and the tags the coils already carry to enter the ERP at zero latency, with the right heat. That is what Connect does in photo-over-analog mode.
The receiving manager photographs the A3 delivery note and every coil tag with the tablet. AI vision extracts heat, gauge, width and supplier, cross-checks them against each other and against the order, and after one validation tap everything is transferred into the ERP — the truck never waited on typing, and heat traceability is intact.
How Connect photo mode works at the receiving dock of a Tier-2 stamping plant:
- A photo of the steelmaker A3 delivery note — AI vision recognizes each steelmaker's delivery note format, without asking them to change theirs and with no predefined templates, and digitizes heat, gauge, width and supplier from the tabular structure by OCR within seconds.
- A photo of every coil tag — OCR of the printed heat code, gauge and width; Connect reads the barcode or QR with the tablet camera, with no dedicated scanner.
- Automatic cross-check against the order — every heat on the delivery note is checked against the coil tags and against the purchase order, without typing a single figure.
- One-tap validation — the receiving manager reviews what was extracted on the tablet; if a heat does not appear on the delivery note, a gauge does not match or a figure falls below the confidence threshold, Connect flags it before the tap. The person decides, not the system.
- Transfer into the ERP over API — after validation, the receipt is recorded in the ERP with coil-heat-lot-part traceability from minute zero: every coil, with its photographed delivery note and its heat, tied to its receipt and ready for the line first-off.
Coil receiving with manual typing vs. receiving digitized with Connect
| Aspect | Classic receiving with manual typing | With iLEAN Connect photo |
|---|---|---|
| Truck time at the dock | 30-45 min, charging demurrage | 5-10 min — photo and one-tap validation |
| Recording the A3 delivery note | Heat, gauge and width typed row by row into the receiving module | A photo of the note — automatic OCR extraction |
| Coil tags | Heat number copied in haste, error-prone | OCR of heat, gauge and width; barcode or QR with the camera |
| Cross-check against the order | Gauges and widths verified by hand against the order | Automatic note-tag-order cross-check |
| Heat errors | Invisible — they surface at the press three weeks later with unexpected properties | Zero — flagged and corrected at the dock |
| Coil-heat-lot-part traceability | Broken — the delivery note gets misfiled between trucks | Complete by photo and API, defensible under each OEM's IATF 16949 requirements |
Impact estimate for your plant — to be validated with your own 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.
- Tier-2 automotive metal stamping plant receiving coils from several steelmakers: a receiving dock, 6-10 coils per truck, dozens of trucks a week, one A3 delivery note with its heats per load.
- Connect photo pilot on receiving — just the dock tablet, with no construction work and no new integration with the steelmaker. It starts after the first kick-off meeting.
- The ERP receiving/warehouse module today requires 3-4 users with an active license to type in receipts. With Connect only 1 validator remains: typical reduction of 50-70% in receiving-module licenses. Estimate to be validated.
- Indicative payback of 3 to 8 months, plus a recurring saving on licenses and the elimination of carrier demurrage at the dock. Estimate to be validated against your data.
- On top of the hard CFO lever — fewer receiving-module licenses, a recurring annual saving defensible in committee, and fewer truck hours charging demurrage — comes the avoidance of the costs of the wrong heat that blows up at the press and of the lost delivery note, measurable from the first month of the pilot. Estimate to be validated.
And the fair question from the plant manager
"What if the system misreads a heat number or ties a coil to the wrong delivery note?" — hallucination is a problem of free generation, not of anchored tasks. In tasks where the AI merely recontextualizes a specific figure from one medium to another (reading the printed A3 delivery note and the coil tags and extracting their fields), the best models brought the error below 1.5% [1]. And even then, the critical call is not made alone: if a heat does not match the delivery note or a figure falls below the confidence threshold, Connect holds the receipt and the person validates with one tap on the tablet. The three safety rings exist precisely for this.
[1] OpenAI paper "Why Language Models Hallucinate", 2025 — on the reliability of AI in anchored tasks.
What people ask about coil receiving in a Tier-2 stamping plant
Why does typing at receiving generate heat errors that do not appear until weeks later?
Because at receiving the manager types fast to release the truck, and a mistyped heat digit does not stand out: the coil enters the warehouse with the wrong identity. Weeks later, when that coil starts the stamping line, it shows up with properties different from the ones expected — yield strength, thickness or a mill certificate that do not match the order — and the problem blows up at the press, not at the dock. With Connect the heat number is read by OCR from the coil tag and cross-checked against the delivery note before the validation tap: the error is corrected at the dock, not on the line three weeks later.
Does the steelmaker have to change their delivery note for Connect to read it?
No, and that is the point. AI vision recognizes each steelmaker's A3 delivery note format with no predefined templates and without asking them to change theirs: it extracts heat, gauge, width and supplier from the tabular structure the same way in every case, however each supplier lays the document out. When a new steelmaker enters the panel, Connect learns their format over the first receipts and the receiving manager's flow does not change: a photo of the delivery note, a photo of each coil tag and one validation tap on the tablet. Neither the APQP agreed with the supplier nor their documentation is touched.
How does the coil tag OCR work?
The receiving manager photographs each coil tag with the tablet and Connect reads the printed heat code, gauge and width by OCR, and reads the barcode or QR with the camera without a dedicated scanner. What is extracted is cross-checked against the A3 delivery note and against the order: if a heat does not appear on the note, a gauge does not match or a figure falls below the confidence threshold, Connect flags it on the tablet before the validation tap. The 6-10 coils on a truck are read and validated at the dock in 5-10 minutes, without typing a single row into the ERP.
Where does the 50-70% saving on receiving-module licenses come from?
Coil receiving today usually requires 3-4 users with an active license for the ERP receiving/warehouse module to type in delivery notes, heats, gauges and widths. With Connect the capture is done by the photo and only 1 validator keeps a license: from 3-4 users to 1, a typical reduction of 50-70% in receiving-module licenses. The calculation is made with the real per-user annual rate in your plant's ERP contract — that produces a recurring annual saving defensible in committee, on top of the truck hours without demurrage and the avoided costs of the wrong heat. Estimate to be validated against your numbers during the assessment.
How is coil-heat-lot-part traceability held together with Connect?
Because the chain is closed at minute zero, not reconstructed afterwards. Every coil is photographed and tied over API to its heat, its lot and its delivery note at the moment of receipt, and that identity travels with it all the way to the stamped part. The coil-heat-lot-part chain stops depending on a delivery note that can get misfiled between trucks: if a seating or suspension Tier-1 customer raises a complaint about a lot of parts, or each OEM's requirements demand evidence of the material origin under IATF 16949, the heat and its mill certificate are recovered from the ERP within seconds and the backwards traceability is complete, by photo.
A Connect pilot at your receiving dock — it starts after the first kick-off meeting.
We work on your plant's real data, not ours. Assessment with no commitment.
Request estimated ROI within 48h ‹ See all 12 metal stamping cases See automotive