Understand the signature
Study how movement changes over time, including the everyday actions that can look like falls.
How it works
Sensors, a hub, agents and an app, working as one
Room sensors describe the routine, the Sharpness Hub is the voice in the room, AI agents plan the next step, and the app shares an approved summary with the people the person at home chooses. Below: our radar research first, then each part of the system and how they connect.
Gen 1 planned for early 2027 · Final capabilities to be confirmed
The sensing technology we are building around
mmWave radar. Movement, breathing and sleep, with nothing to wear.
Radar reads small changes in reflected radio waves: a person crossing a room, a fall, the rise and fall of breathing through the night. Below is our own fall research in motion, then what published studies show radar can sense.
Research in motion77 GHz laboratory research · simulated view
The body is a scripted model fitted to a laboratory recording; its posture is not reconstructed from radar. The trace is simulated from that model.
A brief burst. Then less movement.
In this fitted fall model, movement builds quickly and then settles. We study the shape and timing of that change, rather than treating every strong movement as a fall.
Scaled to each model's own peak. Simulated from the fitted model, not a measured or calibrated signal level.
Model in motion
Research demonstration, not a live detector. Motion alone cannot confirm that someone is on the floor. This 77 GHz laboratory example is separate from the radar sensor planned for Deluxe sleep monitoring, and it is not a demonstrated capability of that hardware.
What published research shows
One kind of sensor, several signals.
Peer-reviewed studies have used radar to track movement, falls, breathing and sleep, with no camera and nothing to wear. What each found, and where StaySharp stands.
Movement
Radio-wave sensing tracked walking speed in people's own homes for up to a year in a 2022 Science Translational Medicine study, and a 2026 millimeter-wave study followed several people moving at once in a large room.
Motion sensors in Premium and Deluxe cover room activity. Radar movement interpretation is a research direction.
Falls
A 2025 review of 80 fall-detection studies found that non-wearable sensors (the family radar belongs to) outperformed wearables, with deep learning giving the best results.
Shown in the demonstration above. Not in the current packages, and no StaySharp accuracy has been established.
Breathing and sleep patterns
A 60 GHz radar beside the bed told wake, REM and non-REM sleep apart with 85 percent agreement against polysomnography in 78 adults (Journal of Sleep Research, 2024).
Radar sleep monitoring is planned for Deluxe.
Sleep apnea
Clinical studies comparing radar-based monitors with overnight polysomnography, one of them in 200 people, found they pick up the breathing pauses of obstructive sleep apnea.
A research direction. StaySharp does not diagnose or treat sleep apnea; talk to a clinician about symptoms.
Where we are building our edge
From radar signals to meaningful context.
Radar sensors themselves are widely available. The advantage we are building is in the interpretation: understanding motion, combining it with household context, and validating what deserves a follow-up.
Add the missing context
Explore range, height and what happens afterward. A motion burst on its own is only part of the story.
Earn confidence through testing
Build toward evaluation across people, rooms and sensor placements. Laboratory models are a starting point; validation in homes is work ahead.
Research direction, not a launch feature. Fall detection is not included in the current packages; it is planned as the research matures, and no accuracy has been established.
The voice connection
Sharpness Hub
The hub is the voice in the room. It runs the daily check-in, speaks reminders, holds a conversation and calls the people the resident trusts, and it connects the room sensors.
- Scheduled check-ins and reminders by voice
- Two-way calls with approved people, and emergency contact calls in every package
- Gen 1 pairs the hub with cloud-based AI agents
- Gen 2 roadmap: AI agents running locally on the hub
The daily context
Room sensors
Small sensors describe a routine without a camera or a wearable: movement in a room, temperature and air quality, and, with the Deluxe radar sensor, sleep monitoring through the night. The radar research above covers what published studies show radar can sense.
- Premium: three motion sensors, an air quality sensor and a temperature sensor
- Deluxe: a radar sensor for sleep monitoring, planned
- Movement in shared rooms and the overnight routine in the bedroom
- Sensor choice and placement reviewed for each home
- Research direction: radar interpretation of movement, breathing and sleep, not a launch feature
The shared picture
StaySharp app
The app brings the daily summary, trends and voice contact together for the people the resident approves. The Sharpness score is the summary's headline, and it is still being validated.
- A daily summary of activity, sleep patterns and outings
- Sharpness score updates, planned for Premium and Deluxe
- Sharing approved by the resident, for family members and care teams
- Contact and alert workflows agreed and tested with each home
The coordination
AI agents turn a routine into a next step
The agents do the planning: a reminder at the right time, a check-in, a conversation, or a prompt for the agreed person to call. Gen 1 runs them in the cloud; Gen 2 plans to run them on the hub.
Learn the daily pattern
Build a picture of the usual day from the sensor signals and the routines the household agrees.
Keep track of what is next
Deliver the medicine, appointment and meal reminders the household or care team set up.
Run the check-in
Ask how the day is going, hold a conversation, and log the answer for the agreed follow-up.
Bring a person in
Route a request for help, a missed check-in or a changed routine to the person named in the response plan.
Planned connections
Fit into the support people already use
App and API connections are on the roadmap. Device compatibility, data access and responsibilities are confirmed before any integration is offered to a household or partner.
Smart pill bottles and home devices
Planned integrations could add device events to reminders and daily summaries. A bottle opening cannot prove a dose was taken, identify an unknown medicine or replace an agreed medication schedule.
Care team software and APIs
Planned interfaces could bring approved summaries into a partner's own workflow. API availability, EHR connections and emergency service integrations are not confirmed launch capabilities.
Start with consent and a clear data boundary
The standard hardware has no camera. The voice design handles interactions started with a wake phrase and does not record ordinary room conversation. Gen 1 uses cloud agents; local agent processing is a Gen 2 goal. Participating households review access and privacy choices before setup.