Why RFID tags fail on metal and liquid

Metal: the antenna detunes

UHF energy reflects off metal, and the metal changes the antenna's resonant frequency. A label that read at 3 m on a cardboard box may read at 10 cm once stuck to a steel panel — or not at all.

The fix: an on-metal label with an absorber layer between the antenna and the metal surface. It restores resonance. Expect 30–100% higher unit cost than a standard label, and it is the only reliable answer.

Liquid: the water absorbs the energy

Water has a high dielectric constant and absorbs UHF energy. A standard label on a full bottle can lose almost all of its range; on an empty bottle it looks fine. This is why testing on empty packaging is the single most common pilot mistake.

Three fixes, in order of cost:

1. Spacer layer — lift the antenna 1–2 mm off the surface with foam. Cheapest, adds thickness.
2. Switch to HF / NFC — 13.56 MHz tolerates liquid far better, at the cost of range (centimetres instead of metres).
3. Custom antenna — redesign for the actual dielectric. Best performance, needs a factory that can iterate on antenna design.

The test that tells you which you need

Put the same label on an empty container and on a filled one. Measure read range at 0°, 45° and 90°.

• Drop under 20% — acceptable, keep the design
• Drop 20–40% — tune the reader power or reposition the label
• Drop over 40% — the design is wrong. A stronger reader will not fix it.

A note on metal + liquid together

A metal can full of liquid is the hardest case in RFID. It usually needs a thick absorber, a raised antenna, or a switch to HF. Plan for a longer sampling cycle — and test on the real, filled, sealed product, never a mock-up.