Barcode & RFID Tracking

Every inventory and traceability system depends on one unglamorous thing: somebody, somewhere, has to record that a physical object moved. Do it with a pen and a register and the data arrives late, incomplete and partly invented. Do it with a scan and it arrives instantly and correctly. Barcode and RFID are simply two ways of making that scan possible, and the choice between them is a practical one, not an ideological one.

What a fabric roll barcode actually has to carry

The most common mistake in a first barcode implementation is to encode the item code. A barcode reading "Cotton Single Jersey 180 GSM Navy" is almost useless, because that is the one thing anyone can already tell by looking.

The barcode should carry a unique roll identifier, and that identifier should resolve — in the ERP — to everything the roll actually is:

The physical label carries the ID; the ERP carries the meaning. That separation matters because roll attributes change — a roll gets re-inspected, gets split, gets partially issued — and a label that encodes the data itself goes stale the moment anything happens to the roll.

Bundle tickets in the sewing room

In garment manufacturing the equivalent object is the bundle: a tied stack of cut panels, one size, one colour, one lay, typically twenty to thirty pieces, with a ticket attached. The bundle ticket is the oldest tracking device in apparel and it long predates computers.

Barcoding it changes what it can do. A scanned bundle ticket at each sewing operation gives live work-in-progress by operation, throughput per operator, and the exact position of an order inside the line — the foundation of [shop floor control](/glossary/shop-floor-control) and of any real [SAM](/glossary/sam-smv)-based efficiency measurement. The ticket also carries the roll IDs the bundle was cut from, which is the link that lets a shade complaint on a finished garment be traced back to a dye batch.

Bundle tickets are usually printed with a barcode per operation — a strip of coupons the operator tears off and the supervisor scans, or a single ticket scanned repeatedly. The former survives factories where scanning discipline is weak, because the physical coupon is evidence; the latter is cleaner where operators are trained and the scan is genuinely done.

Where RFID differs, and where it earns its cost

The functional difference is that a barcode must be seen individually and RFID need not. An RFID reader can read many tags at once, through packaging, without line of sight and without the object being handled.

That bulk-read capability is what changes the economics, and it changes them in specific places rather than everywhere. Counting a fabric store roll by roll with a barcode scanner takes as long as there are rolls; walking the aisle with an RFID reader can capture them together. Verifying that a carton contains the right size ratio means opening it if the garments are barcoded, and does not if they are RFID-tagged. Receiving a shipment can be a gate read rather than an unpacking exercise.

Where RFID rarely pays in a mill is on the process floor itself. Wet processing, heat and mechanical stress are hostile to tags, per-tag cost matters when the object is a low-value bundle, and a barcode on a bundle ticket costs almost nothing and does the same job.

The strong case for garment RFID has mostly come from the retail end of the chain, where item-level tagging supports store stock accuracy and loss prevention — which means a garment factory's RFID programme is often driven by the buyer's requirement rather than the factory's own return. That is a legitimate reason to implement it; it is just a different reason, and it should be costed as a customer requirement rather than as an internal efficiency project.

Getting the implementation right

Three things decide whether a scanning programme survives its first year. Scan at the moment of the physical event, not afterwards from a list — a scan recorded at the end of the shift from memory is a manual entry with extra steps. Scan at the point where the identity is created, so the roll is labelled at doff and the bundle at cutting, before either can go anywhere unrecorded. And make the scan the only way the transaction can be recorded, because as long as there is a manual override, under pressure the manual override becomes the process.

The failure mode is always the same: scanning is introduced alongside the paper system rather than in place of it, both are maintained for a while, the paper one is faster, and within three months the scanners are in a drawer.

Vastra ERP generates roll, bundle and carton labels with identifiers that resolve to lot, shade, length, width, grade and location, and records movements by scan across the [inventory](/features/inventory-management) and [shop-floor](/features/shop-floor) modules, so stock position and work-in-progress reflect physical events rather than end-of-shift reconstruction.