Choosing an RFID tag: surface, distance, environment

Rank the reasons RFID projects fail and the wrong tag sits at the top of the list. The reader is fine, the software works, but the tags do not read. The cause usually fits in one sentence: the tag was not chosen for the surface it is mounted on.

Surface comes first

Radio waves reflect off metal and get absorbed by liquid. Those two materials are the natural enemies of UHF RFID. Stick a standard label straight onto metal and it will not read. Leave half a centimetre of air and performance still swings.

The answer is a tag designed to work on metal, which uses the metal as part of its antenna. Some of the industrial tags we manufacture are built for exactly this, in millimetre-scale housings. For cardboard and textiles a thin, flexible, inexpensive label is enough.

Distance and angle

Required read distance sets antenna size. Small tag, small antenna, short range. You cannot argue with that one, it is physics.

Angle matters as much as distance. A tag facing the reader antenna squarely gives the best result, and items sitting at random on a shelf never guarantee that. So you either use a circularly polarised antenna or mount the tag in a fixed orientation on the product. This is why hanger direction gets discussed in textiles, and position inside the box in jewellery.

Environment

What limits a tag lifetime is not reading, it is where it lives. The questions to ask: what temperature range, will it be washed, will it meet chemicals, will it sit in sunlight, how many years does it need to last.

• Hotel and hospital linen: must survive hundreds of wash cycles and ironing temperatures.
• Cold chain: adhesive behaviour at low temperature is its own problem, and a label tested at room temperature will fall off in a freezer.
• Outdoor and fixed assets: UV resistance and impact protection come first.
• Production lines: oil, dust and abrasion wear the tag out physically.

Volume and cost

Unit tag price becomes the largest line in the budget as volume grows. Tagging five thousand assets a year makes a few cents irrelevant. Tagging two million items a year makes the same few cents decisive.

The trap is dropping to a cheaper tag and losing read rate. One percent of missed reads across two million items is twenty thousand missing records. Whatever you saved on tags, you give back in repeated counts.

Do not decide without a site test

The right method is simple. Take two or three candidate tags, apply them to the real product, on the real surface, in the real packaging. Then read them where the work happens: on the actual shelf, in a full carton, surrounded by other tags. Not on a lab bench.

Do not measure average read distance. Measure how many of a hundred tags read on the first pass. The gap between ninety-eight and a hundred is the only gap daily operations will feel.

We run this test at the start of a project and share the numbers with the customer. When tag selection is guesswork, the mistake gets paid for later by relabelling thousands of items.