Product · May 14, 2026 · 4 min readAll posts →
Product

Barcode scanning built for real-world labels

Pristine barcodes are easy. Warehouses don't have pristine barcodes. Here's how Nautilus scanning is built for scuffed, partial, and double-labeled items.

Any scanner can read a crisp barcode under good light. The barcodes in a real warehouse are scuffed, peeling, frosted with freezer condensation, printed at an angle, or partially torn by a forklift fork. The job isn't reading a perfect label; it's reading the labels you actually have.

Formats, without you choosing one

Nautilus reads the formats warehouses use — Code 128, Code 39, EAN-13, UPC-A, QR, and Data Matrix — in a single pass. The operator never has to tell the app which format they're scanning. It identifies and decodes whatever is in frame.

The scan also feels instant. We hold ourselves to a simple bar: the operator should never register a wait between pointing the camera and seeing the result. A scan that makes someone pause is a scan that breaks their rhythm, and rhythm is most of an operator's speed.

When an item has two barcodes

A common mess: a bag arrives from a supplier with the supplier's barcode already on it, and your own internal label gets added alongside. Now the item has two barcodes encoding two different identities. A scanner that simply grabs whichever code its camera locks onto first will quietly log receipts against the wrong identity.

When Nautilus sees an item that resolves to more than one identity, it asks once which barcode to treat as canonical, then remembers. A wrong scan nobody notices is far more expensive than a one-time question, so we'd rather ask.

Register, locate, and act in one motion

Scanning isn't just lookup. From the scan view an operator can register a brand-new item, find where an existing one lives, pick it, receive it, relocate it, count it, adjust it, or return it. The barcode is the entry point to every action, so the device stays in the operator's hand and the work keeps moving.