Healthcare · Proven

Bathroom and bedside fall detection with sub-second alerting

If a patient falls in the bathroom or by the bed, staff get paged instantly.

Proven Proven live in runtime

Demonstrated on real public research data.

A falling body moves fast, then goes still on the floor. That sudden change alters how WiFi waves bounce around the room in a way nothing else does. The system knows what a fall looks like — even in rooms where cameras are never allowed.

In the physics: A fall is a high-velocity vertical collapse: the reflecting human body produces a broadband micro-Doppler burst (transient energy spread to 5-15Hz) followed by a sudden shift of the body's dominant reflection to floor-level delay/height, distinct from sitting or lying.
Preparing the recorded fieldLoading real LatentField pipeline output…

This is real pipeline output for Healthcare, replayed. The exact reading depends on your room.

Bring this into the real roomPatient Room Field Node3 sensing nodes · ~$140 parts · about 60 min
Go deeper

How the field becomes an answer

Doppler/micro-Doppler transient energy detector RTI differential imaging to confirm body collapse to floor plane PCA+MVE anomaly manifold to reject non-fall bursts (dropped object, door slam)

How we prove it

Fall AUC 0.8 already proven; harden on WiMANS/SenseFi activity classes containing fall vs. sit/lie, then own-hardware simulated-fall trials in a mock bathroom with acoustic/pressure-mat ground truth.

Who it is for

'Private-space fall guardian' — for risk/patient-safety officers eliminating the camera blind spot that drives fall-related liability.

runtime Runs today in LatentField 2.0 as fall_detect

Complete first-room estimatePatient Room Field Node3 nodes · ~$140 parts · 60 min Request kit, subscribe, or set up →
Compatible DIY hardware options
Practitioner Discrete$60–90

15 min · beginner

Operator Single-Venue$280–380

120 min · advanced

← All Healthcare capabilities