Many air waybills never need OCR. When the carrier issues an electronic air waybill, the same data exists as a structured message, and a forwarder connected to it gets pieces, weight and routing without reading a PDF. OCR earns its place on the rest: house bills from smaller forwarders, PDFs forwarded by email, scans attached to an AOG request.
TL;DR. For clearance, extract the AWB number, the HAWB number when the shipment is consolidated, shipper and consignee, airports, pieces, gross and chargeable weight, and the nature of goods. Validate the master AWB number with its check digit, the weight unit, and pieces against the packing list. Doana extracts AWBs and house bills in custom extraction pipelines built for a client's own documents; it is not part of the self-serve product.
MAWB or HAWB: which one does customs need?
Both, for different jobs. The master AWB (MAWB) is the airline's contract with the forwarder. The house AWB (HAWB) is the forwarder's contract with its customer. In a consolidation, one MAWB covers many HAWBs. Customs and security filings increasingly work at house level: in the EU, ICS2 collects entry summary data per house bill for air cargo. Description quality at that level is covered in ICS2 goods description.
For a broker, the HAWB links the goods to the client and the invoice. The MAWB links them to the flight and the arrival. Extract both when both exist.
Which fields matter for clearance?
| Field | Where it is used |
|---|---|
| MAWB number | Arrival, cargo control reference (Canada: the cargo control number follows the carrier code) |
| HAWB number | Client and invoice link in a consolidation |
| Shipper, consignee, notify party | Importer of record and parties check against the invoice |
| Airport of departure and destination, flight | Port of arrival, timing |
| Number of pieces | Against packing list and physical count |
| Gross weight and unit (K or L) | Against packing list; a kg/lb confusion shows up as a factor of 2.2 |
| Chargeable weight | Freight cost, not customs value, but often confused with gross |
| Nature and quantity of goods | Description quality; "consolidation as per attached manifest" on a MAWB |
| Handling information | Sometimes the only place a PO or AOG reference appears |
What errors does OCR make on an AWB?
- Number misreads. An 11-digit MAWB number (3-digit airline prefix plus 8 digits) with one wrong digit. The last digit is the remainder of the 7-digit serial divided by 7, so a misread usually fails the check.
- Weight unit. The unit is a single letter in a small box. Missing it turns 100 lb into 100 kg.
- Pieces vs quantity. Pieces are packages. A box of 20 brackets is 1 piece. OCR is right; a reviewer comparing it with invoice quantity is wrong.
- MAWB taken for HAWB. On a house bill the master number is also printed. Extract both with their labels.
- Consolidation text. On a MAWB, the nature of goods often says "consolidation" and gives no description. The description is on the house bill or the manifest.
How does Doana extract AWBs?
In a custom extraction pipeline. Doana builds it on the client's own AWBs, house bills and related files (invoices, packing lists, T1, certificates of origin, release certificates), with the fields and the CSV layout the client's customs software expects. It starts from the contact form. The self-serve product covers EASA Form 1 box by box and invoice lines in the web app.
Next
Check whether your carriers already send electronic AWB data. Keep OCR for what they don't. For the broker's full invoice and AWB workflow, read Automate commercial invoice and AWB extraction.
Sources
- https://taxation-customs.ec.europa.eu/customs/customs-security/import-control-system-2_en
- https://www.cbsa-asfc.gc.ca/publications/dm-md/pdf/d17-1-10-eng.pdf
- https://www.kuehne-nagel.com/how-to/air-freight/import-control-system-2-ics2
- https://doana.ca/en/contact