Logistics / Supply ChainVerified Real deployment at a named or precisely described company, with a measured result from a primary source.

Cameras at the dock door read every pallet and stop the wrong one before it is loaded (Yazaki, Danfoss, a flooring manufacturer)

Added 2026-10-26 · Last updated 2026-10-26

Manufacturers in this case: Danfoss · Yazaki · all manufacturers using AI

Role

Logistics (shipping and receiving: pallet verification at the loading dock) · Typical users: shipping supervisor, warehouse or logistics manager, forklift driver and dock operator, customer service, continuous improvement or IT lead

The problem

At the loading dock, pallets are checked with handheld scanners and paperwork, under time pressure, and many boxes look alike. A pallet on the wrong truck means a return trip, a late delivery and sometimes air freight, and when a customer reports a shortage or damage, there is no proof of what was shipped. Danfoss estimated that about 600 pallets or cartons a year were loaded on the wrong vehicle at its main warehouse in Rødekro, Denmark; in one case a pallet destined for Italy went to Iceland. At Yazaki, a wiring-harness supplier to car makers, inbound and outbound pallets were scanned by hand, which was slow, error-prone and delayed inventory data in a just-in-time supply chain.

How AI helps

Cameras mounted at each dock door capture every pallet as the forklift passes. Computer vision finds and reads the labels (part numbers, supplier codes, quantities, serial numbers, including AIAG standard labels), can count visible cases and spot damage, and compares the result with the order or shipment in the ERP or WMS. A wrong, unexpected or early pallet triggers a stop signal or an exception to the shift supervisor, and an image of every pallet is stored as proof of what was loaded. People handle the exceptions; the system does the routine scanning.

Step by step

1. Measure the baseline: wrong, missing or damaged pallets reported by customers and carriers, claims paid, re-deliveries and premium freight over the last 6 to 12 months.
2. Check the labels: one readable barcode or standard label (such as AIAG or GS1) on each pallet face that the camera will see; Danfoss had to produce new labels.
3. Decide what the system checks against: the order, the load plan or the delivery date, and what happens on a mismatch (stop light, supervisor alert, block in the WMS).
4. Pilot one dock door, as Danfoss did, and compare the camera's reads with manual scans for a few weeks.
5. Store the pallet images with the shipment number so customer service can answer shortage and damage claims.
6. Extend to all doors and to receiving once the read rate and the exception workflow are stable; review exceptions weekly to fix their root causes upstream.

Tools

• Kargo (computer vision towers at the dock door, integrated with SAP or the WMS): used at Yazaki
• Zetes ZetesMedea loading verification with ImageID machine vision: used at Danfoss
• Accella AI Dock Check (deep learning on LUCID cameras, with Ignition for operator screens): used at the flooring manufacturer
• Other tools to evaluate: the dock and loading verification options of your WMS or label-printing supplier
• Any LLM (ChatGPT, Claude, Copilot, Gemini) with the prompt below, to analyse your past shipping complaints and write the requirements for a pilot

In the AI tools index: Accella AI Dock Check · Kargo · Zetes ZetesMedea · all AI tools used by manufacturers

As reported by the source; most figures are vendor or self-reported and not independently verified.

Human review: cautions

• The camera only checks what the labels and the order data say: wrong labels or wrong orders in the ERP still produce wrong shipments, so review exceptions to fix the cause upstream.
• Danfoss's system is machine vision that reads barcoded labels; the Yazaki and flooring systems use AI computer vision that also reads printed text, counts cases and spots damage.
• Cameras at the dock also film people: tell employees and drivers what is recorded and why, and set a retention period.
• The figures are published by the vendors (Kargo, Zetes) and by the camera maker for Accella; the Danfoss result covers the first 10 months.

You are an experienced shipping and warehouse manager.

I will paste: (1) our shipping complaints and claims for the last 6 to 12 months [date, customer, order, problem reported: wrong item, missing pallet, wrong quantity, damage, wrong destination, early or late; cost if known]; (2) a description of how we pick, stage, check and load pallets today; (3) a sample of our pallet labels as text [fields and barcode types].

Do this:
1. Group the complaints by type and by step where the error most likely happened (picking, labelling, staging, loading, paperwork), with counts, and say which groups a check at the dock door would have caught.
2. List the data a dock camera system would need to compare each pallet with (order, load plan, delivery date), and where it lives in our systems according to my description.
3. Check the sample labels: is there one readable code with the part number, quantity and shipment or order number? List the gaps.
4. Write the requirements for a one-door pilot: what is checked, what happens on a mismatch, who is alerted, what images are stored and for how long, and how we will measure success.
5. List the questions to ask vendors.

Write counts as a spreadsheet formula or table I can check.
Use only the data I paste in; if something is missing, write 'not found' or list it as a question instead of guessing. Do not send, order, post or change anything in any system: produce drafts for my review and approval.

Related cases

• Inventory drones count the plant warehouse on a regular schedule, so inventory is right without stopping production (LAPP USA, GKN Automotive, Hyundai Glovis)
• AI reads supplier delivery notes and prepares the goods receipt in the ERP (Siemens, RAUMEDIC)
• AI-built load planning tool fills trailers better and cuts outbound freight 12% at a trailer maker
• AI moves fast-selling items closer to the pickers and shortens pick routes in the warehouse (Coram, Gildemeister)

All 12 Logistics / Supply Chain use cases →

Sources

• https://www.kargo.ai/blog/customer-spotlight-yazaki
• https://www.zetes.com/en/warehouse-solutions/direct-store-delivery-software/from-600-to-zero-shipping-errors-and-roi-in-one
• https://thinklucid.com/case-studies/ai-vision-system-verifies-palletized-shipments/