Supply chain tracking: NFC, UHF RFID or BLE?

The question almost always arrives in the same form: is NFC or RFID better? It is the wrong question, because it assumes there is one technology that is better in absolute terms. There isn't. There is a technology that matches the gesture people will actually perform.

Three technologies, three gestures

NFC (13.56 MHz). Reading on contact or within a few centimetres, one object at a time, with the smartphone the operator already has in their pocket. The gesture is: I bring my phone close to this specific object because I want to know something about it. Inspecting a piece of PPE, verifying the authenticity of a garment, accessing a machine's documentation. Passive chip, no battery, practically unlimited lifespan.

UHF RFID (860-960 MHz). Reading at a distance, up to several metres, hundreds of tags simultaneously. The gesture is: these objects pass through here and I want to know which ones they are, without touching them. Gates, tunnels, warehouse inventories, shipment checks. It requires fixed or handheld readers and antennas, and a radio design of the reading point that is far from trivial: metals and liquids absorb and reflect.

BLE (2.4 GHz). The tag is active, has a battery and transmits on its own. The gesture is: I want to know where this object is right now, without it passing through anywhere. Real-time location, temperature or impact monitoring, high-value assets moving across large areas. A lifespan of years, not decades.

The decision matrix

| | NFC | UHF RFID | BLE | |---|---|---|---| | Range | contact | metres | tens of metres | | Objects per read | one | hundreds | one, continuously | | Reader | smartphone | dedicated | smartphone or gateway | | Battery in the tag | no | no | yes | | Cost per tag | low | low | medium-high |

The row that almost always decides is the third. If the project requires dedicated hardware in the hands of the people operating it, the cost of adoption grows by an order of magnitude compared with the cost of the tags — and with it the likelihood that the system will be bypassed.

Why the answer is often "more than one"

A real supply chain crosses different gestures. The garment is serialized in production with a textile tag, read in bulk via UHF at the warehouse gate, verified via NFC by the end customer in the boutique. These are not three projects: it is one project with three reading points, and the correct choice is often a dual-frequency tag or two separate tags on the same object.

This is the structural reason why a single-technology supplier tends to propose the technology it sells. A B2B multicommerce that carries NFC, UHF and BLE in its catalogue does not have the same incentive, and can say that for that reading point the answer is "none of the three, a printed code is what you need".

The test to run before signing

Before ordering a hundred thousand tags, order fifty and try them in the exact spot where they will be read, with the real object, with the real people. The read rate in the lab and the read rate on a pallet of cans are different numbers, and no data sheet will tell you by how much.