Fall detection sensors: five kinds, and which room each one suits

What a fall detection sensor can actually measure

Strip the branding away and every sensor in this category measures one of four physical quantities: movement in a field of view, weight on a surface, reflection of a radio or light signal, or acceleration at the body. The fall is never measured directly. It is inferred from a pattern of change, which is why the output of a well-built system is a possible fall that a person confirms.

That single sentence decides most procurement arguments. A sensor that is good at detecting movement in a corridor may be useless behind a drawn curtain; a sensor that works through fabric may not be able to tell a resident from a folded wheelchair. Neither is worse in general. They are worse at different jobs.

The five kinds of fall detection sensor

Five technologies account for almost everything sold into Hong Kong care homes today. They differ in what they need from the room — line of sight, a surface, a radio path, or the resident's cooperation — and that requirement is the real difference between them.

Fall prevention sensors are a different product

In Hong Kong shops the phrase 防跌感應器 is used for two unrelated things: a sensor that notices a fall after it happens, and equipment that reduces the chance of one — a grab bar, a bed rail, a night light, a non-slip mat, a chair alarm. The first is a notification system. The second is the physical environment.

Both belong in a care home, and confusing them in a tender document produces a quotation that answers neither question. A prevention plan starts with the resident's strength, medication review, vision, footwear and the route to the toilet at night. A detection plan starts with the rooms where a fall would otherwise go unnoticed. Write them as two lines in the same budget, not as one line called safety.

Which sensor suits the bedroom, the bathroom and the corridor

The bedroom is where privacy is tightest and where many falls happen, at the moment of getting up or lying back down. A method that needs no camera and no wearable wins here: radar or Wi-Fi sensing, plus a pressure mat if the concern is specifically the exit from bed. A camera in a shared bedroom is a consent problem before it is a technical one.

The bathroom is the hardest room. It is small, wet, often has a closed door, and a resident may be undressed. Radar and Wi-Fi sensing both work through a door and in the dark, which is why they are usually the only serious options once the camera and the wearable have been ruled out on dignity and compliance grounds.

The corridor is the easiest and the least valuable on its own. Passive infrared tells you someone walked past; it does not tell you they did not come back. A corridor sensor is a useful complement to a room sensor, and a poor substitute for one.

What no sensor will tell you

No sensor tells you whether the resident is hurt, whether they hit their head, or whether they can be moved. It tells you a posture changed in a way that resembles a fall, and the rest is clinical judgement made by a person in the room. Copy that promises more than that is describing a product nobody has built.

No sensor catches every fall either. A resident who slides gently down a wall, or who is lowered to the floor by a carer, may never produce the pattern the model looks for. This is not a reason to skip sensing; it is a reason to run sensing alongside the checks a good night shift already does.

And every sensor produces alerts that turn out to be nothing. The measure that matters is not the alert count but the interval between alert and a staff member at the bedside — which is why Elyvo is built around that interval rather than around a sensor list.

Questions to ask before you buy

Ask what the sensor physically measures, and ask the vendor to say it without the word accurate. Ask which rooms it is specified for and which it is not. Ask what happens when the resident is behind a curtain, under a blanket, or in the shower. Ask where the alert lands, who acknowledges it, and what record remains afterwards.

Then ask about the building. Radar and Wi-Fi sensing both depend on installation: node placement, ceiling height, wall material and the number of rooms per floor all change the result. A pilot in two rooms is worth more than a specification sheet, and a pilot that records alert-to-arrival time is worth more than one that records satisfaction.

For the camera option specifically, our camera fall detection piece covers the consent questions; the SteadiCore home page shows how the sensing methods sit inside one response loop, and bed exit alarms covers the night-time leaving problem in detail.

If the decision sits with a family rather than a facility, the same group writes about caregiving coordination at Easikin (opens in a new tab) and about the documents an ageing parent needs at Legaris (opens in a new tab).

Questions people ask

Which fall detection sensor is the most accurate?

There is no measured ranking we can quote across technologies, and accuracy depends on the room. A wearable is precise when it is worn and useless when it is on the bedside table; Wi-Fi sensing keeps working when nothing is worn but resolves posture less finely.

Can one sensor cover a whole flat or a whole floor?

Not usefully. Signal-based methods degrade with distance and walls, and a single device covering several rooms cannot tell you which room the event happened in, which is the first thing staff need to know.

Do we still need a wearable if the room is sensed?

Only if the resident leaves the building or walks in unsensed areas. Inside a sensed room the wearable adds little and costs the resident comfort, charging and a device to remember.

What does 防跌感應器 mean in a Hong Kong shop?

Usually a fall-prevention device rather than a fall detector: a mat, a rail, a night light or a chair alarm. Check what the product measures before comparing prices.