Zero UI · Passive · Dignified Sensing

The home that
looks after you
without watching you

Six invisible sensing technologies woven into the fabric of the home. No cameras. No wearables. Nothing to remember or charge. Each one detects a single meaningful signal — and only that signal — to help an older person stay safe, independent and connected on their own terms.

No cameras No wearables 🔒 Processing stays in the home ✋ Person owns all data ↩ Always reversible
🏡
Invisible by design
Devices blend into the home. Sensing, not surveillance.
🧠
Insight, not data
All raw signals are processed locally. Only derived scores are ever transmitted.
🌱
Graduated choice
One device at a time, chosen freely, added in response to what the person says matters most.
Always reversible
Any device can be removed at any time. The person is always in control.
The sensing stack

What each device does for you

Filter by what matters most — to you, or someone you care for.

📡
Sensor 01
Presence & Movement Radar
mmWave radar · ceiling mount · no image

"Knows you fell before you can call for help — without ever seeing you."

A small disc fitted to the ceiling. It emits an invisible radio wave — the same technology used in automotive safety systems — and detects movement, breathing and posture from the way that wave bounces back. There is no camera, no microphone, no image of any kind. It recognises patterns: a normal walk to the bathroom at night, a fall, a person lying still who shouldn't be.

What it detects
  • Falls and sudden posture changes — no button to press, no delay
  • Gait speed and walking steadiness — the single most reliable frailty marker
  • Breathing rate during sleep — changes can signal infection or cardiac strain early
  • Night-time movement — restlessness, frequent bathroom trips, stillness for too long
🎙
Sensor 02
Acoustic Intelligence
Sound pattern AI · edge processed · never recorded

"Hears that the kitchen is busy — never what is being said."

This sensor listens to the acoustic texture of a home — cooking activity, conversation, the television, silence — and classifies patterns without ever recording speech. All audio is processed on the device and immediately discarded. What Assistiv receives is a score: "conversational activity was low this week." Words are never heard, stored or transmitted.

What it detects
  • Duration and frequency of conversation — an early signal of social withdrawal
  • Kitchen activity complexity — meal preparation declining in scope is a meaningful marker
  • Periods of sustained silence — relevant to isolation and wellbeing
  • Unusual sound events — a fall, extended coughing, a cry
Sensor 03
Home Energy Intelligence
Smart energy monitor · consumer unit clamp · no room entry

"Knows the kettle was boiled at 7am — the simplest sign that the morning is going well."

A clip-on sensor attached to the home's electricity supply reads the unique energy signature of every appliance — kettle, microwave, oven, television — and learns to distinguish them without entering any room. This is the least intrusive sensor in the stack: nothing in the living space, no camera, no microphone. It reads electricity, not behaviour directly.

What it detects
  • Morning routine completion — kettle and TV use at expected times
  • Meal preparation activity — microwave and oven use as proxies for eating properly
  • Appliance safety concerns — items left on longer than expected
  • Changes to daily rhythm — later wake times, skipped routines, shifted patterns
🚪
Sensor 04
Door & Room Contacts
LoRaWAN contacts · battery powered · discreet fit

"Knows the front door opened at 10am — the walk to the corner shop happened."

Small magnetic sensors fitted to doors, the medicine cabinet or the fridge. Each records a simple event: opened or closed, and when. A medication cabinet opened at the right time each day means tablets are being taken. A front door not opened for several days may indicate isolation or illness. No technology enters any room, no camera, no microphone.

What it detects
  • Medication adherence — cabinet opened at the expected time each day
  • Social and outdoor activity — front door opened, daily rhythm maintained
  • Fridge access — an indirect but reliable indicator of eating and routine
  • Wandering risk — unexpected door activity at night
🌙
Sensor 05
Sleep Intelligence Pad
Under-mattress · completely invisible · no wearable

"Understands rest the way a sleep clinic would — from beneath the mattress, invisibly."

A thin flat sensor that slides under the mattress and disappears. It detects movement, heart rate variability and breathing rhythm through the mattress itself, identifying sleep stages without any wearable device — no watch, no band, nothing to put on or remember. Poor sleep is both an early warning sign and an accelerant of frailty. This closes a gap for bedrooms where a ceiling device would feel unwelcome.

What it detects
  • Sleep duration and continuity — broken sleep is an early frailty and cognitive signal
  • Heart rate variability during rest — changes associated with infection or cardiac strain
  • Bed and rise times — drift in these patterns is a meaningful cognitive and mood marker
  • Restlessness patterns — relevant to pain, anxiety and respiratory health
🌡
Sensor 06
Home Environment Monitor
Temperature · air quality · humidity · passive

"The room that gets too cold in January — quietly raising the risk of a fall."

A small sensor, about the size of a smoke alarm, monitoring air quality, temperature and humidity. Cold rooms increase fall risk and cardiovascular strain; poor air quality adds cognitive load; damp accelerates respiratory disease. This sensor asks nothing of the person and requires no interaction. It simply notes whether the environment someone is living in is safe.

What it detects
  • Room temperature — hypothermia risk in winter, particularly for those managing fuel costs
  • Carbon dioxide levels — elevated CO₂ is associated with reduced concentration and sharpness
  • Humidity — damp environments accelerate respiratory conditions and fall risk
  • Thermal comfort trends — patterns suggesting heating is being avoided
🚽
Sensor 07
Toilet Flush Sensor
LoRaWAN contact · cistern mount · battery powered

"The single most clinically valuable sensor in the home — and the least discussed."

A small contact fitted to the toilet cistern records each flush — nothing more. From that simple signal, the pattern of use across the day and night becomes a remarkably precise clinical proxy. UTI is the leading cause of acute hospital admission in older women, and the warning signs are written in bathroom frequency long before symptoms become severe enough to call for help.

What it detects
  • Frequency and timing of toilet use — elevated urgency or frequency is an early UTI signal
  • Nocturia patterns — multiple night-time flushes indicate sleep disruption, bladder or prostate issues
  • Reduced frequency — a marker of dehydration, constipation or declining fluid intake
  • Absence of morning use — alongside other sensors, a meaningful wellbeing flag
💧
Sensor 08
Water Flow Monitor
Pipe clamp · under-sink · no plumbing changes

"Knows whether the tap ran this morning — a quiet proxy for drinking, washing and self-care."

A non-invasive clamp on the cold water pipe under the kitchen sink detects flow without any plumbing changes. When the tap ran, for how long, and at what times of day becomes a composite signal for hydration, washing up and morning self-care routine. Combined with the kettle signal from energy disaggregation, it builds a complete picture of fluid intake — a direct driver of UTI risk, cognitive function and frailty progression.

What it detects
  • Tap use frequency and duration — proxy for drinking water and hydration across the day
  • Morning washing-up activity — an indirect but consistent self-care marker
  • Absence of flow — days with very low tap use signal dehydration risk
  • Pattern shifts — reduced water use correlating with declining activity or mood
🪜
Sensor 09
Stair Activity Sensor
Pressure mat · under carpet · completely hidden

"When someone stops using the stairs, it's rarely a decision — it's a signal."

A thin pressure mat embedded under the carpet on the bottom stair tread. Invisible once fitted, it counts stair ascents and descents across the day. Stair use is a validated proxy for physical capability, confidence and lower-limb strength — all primary frailty markers. The point at which someone begins avoiding the upstairs is a meaningful clinical threshold, reached gradually and silently, often weeks before it becomes visible to family or clinician.

What it detects
  • Daily stair use frequency — declining trips upstairs is an early physical capability marker
  • Time-of-day patterns — avoiding stairs in the evening may indicate fatigue or pain
  • Sitting on the stairs — a distinct pressure pattern indicating a rest stop or post-fall position
  • Sudden cessation — a sharp drop in stair use often precedes formal frailty identification by weeks
💨
Sensor 10
Continuous CO Monitor
Sub-threshold sensing · continuous · LoRaWAN

"A standard CO alarm sounds at danger. This sensor acts weeks before danger arrives."

Standard carbon monoxide alarms prevent acute poisoning — they trigger at already-dangerous levels. A continuous low-level CO monitor operates at an entirely different threshold, detecting chronic sub-acute exposure from a poorly ventilating boiler or gas ring left on low. At 5–15ppm — well below alarm thresholds — CO causes cognitive symptoms frequently misread as dementia, depression or simple confusion. Catching and resolving the source can reverse apparent cognitive decline.

What it detects
  • Chronic low-level CO (5–15ppm) — causes fatigue, headache and confusion often attributed to cognitive decline
  • Gradual boiler efficiency changes — a slowly rising baseline indicates developing flue or combustion issues
  • Cooking ventilation patterns — CO spikes after cooking indicating inadequate kitchen extraction
  • Appliance safety — gas rings partially left on, a significant risk in early cognitive decline
🧪
Sensor 11
Odour Intelligence Sensor
NH₃ + H₂S · LoRaWAN · electrochemical · bathroom or kitchen

"Detects what a home visit would notice — before the home visit happens."

A discreet sensor measuring ammonia (NH₃) and hydrogen sulphide (H₂S) continuously. Ammonia spikes are a sensitive proxy for incontinence and personal hygiene neglect — early markers of self-neglect in the Missing Middle. H₂S elevation indicates organic decay: food left out, bins not emptied — mapping directly to the ASCOT domain of Food and Drink and reflecting declining executive function. Neither signal is intrusive. Both can trigger an amber alert to the care wallet long before a crisis visit reveals the same picture. Reference device: Milesight GS301 LoRaWAN Bathroom Odour Detector.

What it detects
  • Ammonia (NH₃) spikes — proxy for incontinence, hygiene neglect, or failing personal cleanliness routine
  • Hydrogen sulphide (H₂S) elevation — indicates organic decay, food left out, bins not emptied
  • Persistent odour patterns — distinguishes a single event from a sustained change in home condition
  • ASCOT domain mapping — directly signals Personal Cleanliness & Comfort and Food & Drink decline
Privacy architecture

Raw data never
leaves the four walls.

Every sensor processes data locally — on a small hub inside the home — before anything is transmitted. What leaves is not data. It is insight: a score, a trend, a flag. The original signal is discarded at the edge.

  • All raw sensor data is processed inside the home, on a local edge device. It never touches the internet.
  • Only derived insight scores are transmitted — not raw readings, not audio, not images of any kind.
  • The person holds the key. Family and clinicians see only what the person has chosen to share.
  • Any sensor can be switched off or removed at any time, with no effect on the rest.
  • No cameras. No voice recordings. No biometric identifiers. No data sold to third parties. Ever.
How data flows
📡
Sensor detects a signal
Raw data — radar wave, sound pattern, energy reading — captured locally inside the home.
🏡
Edge processing inside the home
AI model classifies the signal on a local hub. Raw data is discarded immediately. Only the classification remains.
📊
Insight score transmitted
A derived score — "gait stable", "sleep disrupted", "routine maintained" — is sent to Assistiv Cloud.
🔑
Person controls who sees it
Family, GP or care team see only what the person has chosen to share. The person holds the key.
Your pace, your choice

One device at a time,
arrived at freely

No one is handed a box of sensors. Assistiv begins with a conversation. Technology enters the home only in response to something the person says matters to them.

01

A conversation, not a form

Assistiv begins by asking what matters most to you about staying at home. No clinical language. No forms. The priorities you name shape everything that follows.

02

One chosen device

If a sensor would help with something you named — night-time safety, peace of mind for the family, remembering medication — it is offered as a response to your words, not a product recommendation.

🌱 Starts with one
03

In your own time

Each sensor is introduced only when you are ready. Any device can be removed. Assistiv does not need all six to be useful — even one sensor, chosen for the right reason, makes a meaningful difference.

↩ Always reversible