Every room, aware.
The Wi-Fi already in the room detects presence and falls, and alerts care staff. No cameras, no wearables.
In care homes, night shifts are thin and rooms are many.
people over 65 in Italy fell at least once in the past year. In care homes the risk peaks at night, when fewer staff are on the floor and no one is in the room.
Source: Istituto Superiore di Sanità, PASSI d’Argento surveillance 2022–2023
Cameras
cost residents their privacy, and in workplaces they are restricted by Italian labour law.
Wearables
get forgotten, removed or refused.
A third way.
The Wi-Fi already in the room.
Every Wi-Fi signal that crosses a room is distorted by what happens inside it.
A person walking, sitting down, breathing or falling changes how the signal travels, subcarrier by subcarrier. We read those changes.
Channel State Information (CSI): human activity, without a single image.
Wi-Fi signal
Small nodes read how the Wi-Fi signal changes inside the room.
Local analysis
An Edge AI model recognises events such as falls or long absences from bed, on site.
Staff alert
A carer gets the alert and goes to check. A person always makes the call.
Three signals, one goal: noticing in time.
- 01In development
Fall detection
and immediate alerts for care staff
- 02In development
Presence & activity
bed exits and long absences, including at night
- 03Research
Respiratory patterns
as a non-contact wellbeing signal
SilverPulse is not a medical device: it flags events to be checked, it does not diagnose. We will not publish accuracy figures until we have measured them in the field with our partner sites.
Two settings, one need: nobody should be left on the floor without anyone knowing.
Care homes
Night shifts, reduced staff, many rooms to watch. Alerts for falls, bed exits and long absences, with no image of residents.
Lone workers
We detect a “person down” event, not productivity. A safety system, not a surveillance one, introduced through the procedures required by Italian labour law.
A control room for the people who care.
Room status, alerts and event history in a single view, designed for the night shift.
- 03:46Possible fallRoom 205 · waiting to be taken
- 03:31Out of bed for 15 minRoom 208
- 02:58Back in roomRoom 203 · closed by M. R.
Site map
Every room and its status, floor by floor.
Alert centre
Who took it, what happened, and the full history of each event.
Reports
Night-by-night trends for coordinators and management.
Privacy is our starting point.
No images are captured. Data is processed locally.
No cameras, no microphones
We only use the Wi-Fi signal. There is no image of a resident, not even by mistake.
Raw signal stays on site
Analysis runs locally. Only events reach the platform, never the signal data.
Shared DPIA before installing
A data protection impact assessment prepared with your DPO before any installation.
Resident safety, not staff monitoring
The system tracks safety events. It does not measure staff time or activity.
From pilot to every home.
Today we run pilots with dedicated nodes to validate the system in the field. The goal is to run the same software inside routers, with no extra hardware.
Pilots with dedicated nodes
Small Wi-Fi nodes placed next to the site’s network, no construction work. They validate the system and train the models with 3–4 partner sites in Lombardy.
Inside the router
The same software on access points and routers, together with network operators. Nothing extra to install.
In every home
A safety service for people living alone, switched on through the connection they already have.
Programma Titanium
Pre-acceleration programme, Università di Milano-Bicocca.
Giacomo Cerizzi · Nicola Giampietro
Co-founders, supported by a technical team working on Wi-Fi sensing and Edge AI.
Let’s talk.
If you run an elderly care facility, work on workplace safety or on Wi-Fi sensing. Even just for a critical opinion.