A display and a setup wizard are among the most expensive parts of an IoT device — and the user touches them once. NFC in IoT removes that cost: one tap of a phone against the enclosure hands over network credentials, keys and operating parameters before the device even powers up its long-range radio. It solves three problems at once: onboarding, servicing and authentication.
In short: NFC in IoT is short-range radio communication (13.56 MHz, up to roughly 4–10 cm) used mainly for tap-to-pair pairing, screenless device configuration and the authentication of users and service technicians. It is not meant for telemetry — that job is handed over to BLE, Wi-Fi or a cellular network.
What is NFC and how does it work in IoT devices?
NFC (Near Field Communication) is a family of inductive connectivity standards working at 13.56 MHz with data rates of 106, 212 or 424 kbit/s. It relies on magnetic coupling between loop antennas rather than wave propagation — hence the few centimetres of range and the resistance to eavesdropping from a distance.
The foundations are ISO/IEC 14443 (type A/B proximity cards), ISO/IEC 15693 (vicinity tags) and ISO/IEC 18092 for peer-to-peer mode. The data layer is organised by the NFC Forum NDEF format (NFC Data Exchange Format), which lets any phone read a record from the device without a dedicated app.
Three modes are used in practical design work: reader/writer (the device reads a tag), card emulation (the device presents itself as a card to a reader) and peer-to-peer, today largely superseded by the reader modes. Most IoT products use NTAG-class chips (NTAG 5, NTAG I2C Plus) or ST25DV, which combine the radio interface with an I²C bus to the microcontroller.
NFC or BLE? When to use which technology
NFC and Bluetooth Low Energy do not compete, they complement each other: NFC establishes trust in a single gesture, BLE carries the traffic. The choice depends on whether you need a one-off expression of user intent or a continuous data session.
- Range: NFC 4–10 cm, BLE 10–100 m. The short range of NFC is a security feature, not a limitation.
- Power draw: a passive NFC tag draws 0 µA from the device battery — it is powered by the reader field. That lets you configure or diagnose a product sitting on a warehouse shelf with its battery disconnected.
- Time to connect: NFC pairs in about 100 ms versus the 1–3 s typical of BLE scanning and pairing.
- Throughput: BLE reaches hundreds of kbit/s, NFC realistically 10–30 kbit/s — which is why NFC carries configuration, not a stream of measurements.
If you are designing a mobile app that controls a device, combine both mechanisms — we covered them in our article on Bluetooth Low Energy in IoT.
Tap-to-pair: pairing and configuration without a screen
In tap-to-pair, a phone tap passes the device a complete set of start-up parameters: Wi-Fi SSID and password, MQTT broker address, tenant identifier and session key. It eliminates the device-as-hotspot mode that generates a sizeable share of support tickets.
What does a typical flow look like?
- The installer scans the enclosure with a phone; the NTAG chip returns an NDEF record with the serial number and model identifier.
- The app verifies the device in the cloud and writes an encrypted configuration block into the tag memory.
- The NFC chip raises an interrupt to the microcontroller (a GPIO pin), and the firmware reads the data over I²C and brings up the appropriate radio interface.
- The device registers with the platform and reports its status — from then on the NFC channel serves servicing only.
This pattern fits neatly with the approach described in our post on IoT device provisioning and zero-touch onboarding: NFC supplies physical proof of presence, while the cloud decides on permissions.
Authentication and security: is NFC safe?
NFC on its own provides no confidentiality — a raw NDEF tag can be read by any phone. Security is built one layer up, on cryptography inside the tag chip and in the firmware.
Modern chips offer AES-128 authentication (such as NTAG 424 DNA mechanisms with a dynamic SUN/CMAC code), password protection for reads and writes, and a tap counter that guards against replay. Keys belong outside the microcontroller flash — in a dedicated secure element acting as a hardware root of trust. The complement is secure boot, which guarantees that data from the tag is received only by trusted firmware.
The practical rules we apply: configuration delivered over NFC has a short validity window (60–300 s), is signed and bound to the serial number, and the user memory area is locked once deployment is complete.
Hardware design: antenna, enclosure and integration
An NFC antenna is a loop with an inductance of typically 1–3 µH, tuned with capacitors to resonate at 13.56 MHz. The most common mistake is placing it over a ground plane or next to a metal enclosure — eddy currents can cut the range to zero.
- Leave a cut-out in the copper layers under the loop and a minimum of 3–5 mm clearance from components.
- With metal enclosures, add a ferrite shielding layer between the antenna and the metal.
- Mark the tap point on the enclosure — a 5 cm range means the user has to know where to aim.
We covered these dependencies in more detail in our material on antenna design in IoT devices.
Where does NFC in IoT deliver the most value?
NFC pays off most where a single-touch gesture matters and the device has no user interface. Four mature scenarios stand out:
- Hospitality and access control — cards and phones as keys; we expand on the topic in our piece on smart access control.
- Industrial servicing — reading logs and counters from a device with no power and without dismantling it.
- Manufacturing — programming parameters at the end of the assembly line, inside a closed enclosure.
- Anti-counterfeiting and identification — tags with dynamic authentication as a digital certificate of authenticity.
Frequently asked questions (FAQ)
What is the real-world range of NFC in an IoT device?
The practical range is 1–4 cm for a typical antenna in a plastic enclosure and up to 8–10 cm with a large loop and a strong reader. Datasheet figures assume perfect tuning; enclosure, metal and battery usually cut the distance by 30–50 percent. Always measure on the final production sample.
Does NFC work when the device has no power?
Yes, provided you use a passive tag with non-volatile memory, for example NTAG I2C Plus or ST25DV. The reader field powers the tag chip, so you can read the serial number or firmware version, or write a configuration, with the battery completely flat. Reading data from the microcontroller itself does require the device to be powered.
Will NFC replace Bluetooth Low Energy in my product?
No. NFC handles a one-off, deliberate gesture: pairing, authorisation or configuration. Continuous data exchange, notifications and firmware updates are handled by BLE or Wi-Fi. The optimal design combines both interfaces: NFC establishes trust and passes the keys, BLE carries the working traffic.
Summary and key takeaways
NFC in IoT is not a gadget but a way to lower BOM cost, shorten installation and tighten authentication. Key takeaways: use NFC for intent and keys, not telemetry; encrypt configuration data and give it a time limit; plan the antenna early in the PCB design, because later fixes are expensive.
At FSS Technology we design the whole chain — from the PCB antenna layout, through firmware handling NFC and the secure element, to the backend that registers devices. If you are planning a product with tap-to-pair, or want to simplify the rollout of an existing fleet, see our connected device services and let us talk about your scenario.