NFC vs UHF RFID for inventory
NFC is cheap to start with because the reader is a phone people already own, and it suits equipment scanned deliberately, one item at a time. UHF RFID reads hundreds of tags at several metres in a sweep. The dividing line is not price, it is whether anyone has to count things in bulk.
The line between them
Reading a hundred items one at a time is not a slower version of reading a hundred items at once. It is a different activity, with a different labour cost, and it does not become acceptable by buying faster hardware.
UHF RFID exists to read many tags at once at a distance of metres, which is what GS1 standardises for supply chain use. NFC cannot do this and never will, because at 13.56 MHz the coupling that makes a passive tag work only exists at very short range.
Why NFC looks cheaper than it is
The entry cost is genuinely low. Tags cost cents, the reader is a phone the staff already carry, and a pilot can run on a free application with no procurement at all. For tagging fifty pieces of equipment for maintenance records, this is exactly right and nothing else comes close.
The cost that does not appear in the pilot is time per item. A process that works beautifully at fifty items becomes an hour of somebody walking and tapping at a thousand, every time a count is required. That cost recurs, and it grows with the estate.
The failure pattern
An NFC inventory pilot succeeds, the estate grows, counts start being skipped because they take too long, the data goes stale, and the system is quietly abandoned. Nothing broke. The process simply stopped being worth doing.
What UHF actually costs
Readers, and they are not phones. A handheld UHF reader is a dedicated device with a dedicated price, and fixed portal readers are an installation rather than a purchase.
Tags are comparably cheap, sometimes cheaper in volume. The expense is on the reading side, plus the integration work, plus the tuning that a UHF deployment always needs because reflections and metal make coverage unpredictable in ways NFC never is.
How to decide before spending anything
Count the reads, not the items. If a person will scan items one after another more than a handful of times in a session, you are heading for UHF whatever the item count says.
Then ask whether a phone must be the reader. If the answer is yes because the users are the public or field staff without dedicated hardware, UHF is off the table and the process has to be designed around one tap at a time.
The two questions together resolve almost every case. Deliberate individual reads by people carrying phones: NFC. Bulk counting by people carrying equipment: UHF. Trying to make NFC do the second is the most expensive mistake available here, because it is only discovered after the tags are applied.
Frequently asked questions
Can a phone read UHF RFID with an accessory?
Sled readers that attach to a phone exist and work, which makes the phone the interface rather than the radio. It is a legitimate middle path when staff already carry phones, and it is dedicated hardware with dedicated cost.
Are UHF tags more expensive than NFC tags?
Not necessarily, and in volume they can be cheaper. The cost difference between the two approaches sits in the readers and the integration, not in the tags.
Can I start with NFC and move to UHF later?
Only by retagging everything, since the two are incompatible at the radio layer. If bulk counting is plausible within the life of the deployment, it is cheaper to decide now than to apply tags twice.
Does UHF work on metal and liquids?
Both interfere with it, in different ways from NFC and often more severely at a distance. Tags designed for metal exist on both sides, and a UHF deployment in a metal rich environment needs a survey rather than an assumption.