The Light That Notices You Are Still There: How mmWave Radar Ends the Motion-Sensor Blackout

The light that gives up on you

Anyone who has sat still on a sofa and watched the ceiling light click off has met the real limit of the passive infrared sensor. A PIR does not see people; it sees the change in infrared heat crossing its field of view. Walk in and the moving warm silhouette trips it instantly. Settle into a book, a film or a video call and, within a fixed cool-off window, the sensor decides the room is empty. The fix most households reach for is a longer timeout, which only trades one annoyance for another: lights that hang on for ten minutes in a room nobody is in. The Home Assistant forums are full of the same workaround, a timer padded a few seconds past the sensor's own three-minute cool-off, because a motion sensor asked to behave like a presence sensor can only ever approximate the job.

Radar that reads a breath

Millimetre-wave sensing closes that gap by changing what the room is asked. Instead of waiting for a warm body to move, an mmWave sensor emits a continuous high-frequency chirp — frequency-modulated continuous wave, or FMCW, radar, usually in the 24 GHz or 60 GHz band — and measures the tiny frequency shift in the reflection that comes back. That shift is sensitive enough to register the millimetre rise and fall of a chest as someone breathes. A person reading motionless still reflects a moving signal, so the room stays classed as occupied. The sensor is not a camera and holds no lens or microphone; it resolves the space as a point cloud of distance and micro-motion rather than an image, which is why it can confirm a body in the dark, behind a thin screen or through clothing without recording anything a person would recognise as surveillance.

One room, thirty zones

The more consequential shift is spatial. A PIR gives a single verdict for its whole cone: motion, or nothing. A radar sensor knows where in the room the reflection sits, and that turns a binary trigger into a map. The Aqara FP2, the sensor that pushed this idea into ordinary homes, covers a space of up to about forty square metres and lets the installer draw as many as thirty separate zones inside it, tracking up to five people at once and running the logic locally so the automations keep working with the internet down. Map the sofa, the reading chair and the kitchen island as distinct zones and each can hold its own light: the reading corner brightens when you sink into it and dims when you leave, with no wave of the arm and no dependence on anyone moving. Lighting stops being triggered by the act of entering and starts following where a body actually is.

Why an interoperable standard changes it

Until recently this precision was a single-vendor trick, locked to one app and one radar module. What makes it an industry direction rather than a gadget is standardisation. Matter 1.4 extended the standard's occupancy model to cover radar, vision and ambient sensing, with adjustable sensitivity and event-based reporting, so a presence reading can be described in a common language every certified hub understands. Newer sensors carry it further: radar presence and ambient-light data now travel over Matter-over-Thread into Apple Home, Alexa, Google Home, SmartThings and Home Assistant alike, and certified sensors can report several people across a set of zones without any visible motion. Aqara's own successor to the FP2, the spatial sensor built for Matter, tracks the position and posture of as many as ten people across roughly a ten-by-eight-metre area. When presence is a standard signal rather than a proprietary one, a designer can specify the sensing layer without marrying the whole house to a single brand.

Designing light around presence

For anyone planning a scheme, the practical move is to stop thinking in switches and start thinking in occupied zones. Pair the radar's built-in light sensor with its zones so a lamp only rises when the daylight in that specific corner has genuinely dropped, not because a clock said evening. Reserve a fast PIR, or the radar's own approach detection, for the doorway where instant response matters, and let mmWave hold the deeper, sit-still parts of the room. Because the sensor reports distance, a hallway can light in graded steps ahead of a walker rather than flooding on all at once. The result is the quiet ideal of good lighting automation — a room that is simply lit correctly whenever someone is in it, and dark within moments of the last person leaving, without anyone having learned where a switch is.

The limits worth naming

Radar is not magic, and pretending otherwise leads to bad installs. Its sensitivity to micro-motion is also a sensitivity to the wrong motion: a fan, a fluttering curtain or a pet can register, and a beam that passes through a stud wall may catch a person in the next room unless the zones are drawn to exclude it. The better sensors answer this with careful zone editing and, increasingly, a hybrid design that wakes on PIR and confirms with radar, which is also how battery models reach a multi-year life instead of demanding a wire. mmWave sensors cost more than a PIR puck, and the mains-powered ones need a spur. But for the rooms where people stop and stay — the sofa, the desk, the bath, the bed — presence sensing finally lets the light do the one thing the motion sensor never could: notice that you are still there.