Privacy-first room safety sensing to reduce time-to-check
Non-contact detection and instant alerts to staff devices. No cameras. No audio. On-device processing by default.
- GDPR-aligned principles
- No cameras
- Works offline; buffers alerts*
*Behavior depends on deployment; see Spec snapshot.
Not for diagnosis or prevention. Not a medical device.
- Install time
- ~30 mins/room*
- Sensing
- Non-contact, no video/audio
- Processing
- On-device first
- Alerts
- Mobile & central tablet
*Indicative; site conditions vary.
The problem
Unwitnessed events and long-lies drive harm, complaints, and cost. Night shifts are thinly staffed; cameras are unacceptable for resident dignity.
Our approach
Continuous, privacy-preserving sensing converts movement patterns into actionable alerts. Staff see what’s urgent, not a feed of people’s rooms.
Non-contact signals detect occupancy patterns, bed-exit, and irregular movement—without identity or video.
On-device models flag suspected falls or prolonged inactivity with tunable thresholds to cut noise.
Notifications route to the right carers on mobile or a central tablet for quick checks.
Spec snapshot
- Wall/ceiling mount; mains or PoE*
- Setup via QR; zero-touch updates
- Local gateway optional
*Power option depends on site.
- Suspected fall / inactivity / night-wandering
- Quiet hours and per-room tuning
- Mobile push + tablet dashboard
- No raw video or audio collected
- Signals summarized; retention configurable
- On-device first; cloud optional
- Standalone or bridged to existing alerts
- CSV/JSON export for BI
- Nurse-call interoperability on roadmap
- TLS in transit; device storage encrypted where applicable
- Role-based access; per-user accounts; audit trails retained 90 days by default
- Data-minimal: event summaries by default; retention 30 days by default (configurable)
- Works without Wi-Fi; queues alerts for delivery*
- Facility-wide health and per-room status
- Handles multiple occupants in a room without identification
*Delivery depends on gateway/coverage.
- Alerts are based on activity patterns; they may be false or missed.
- No diagnostics, vitals, or clinical decisions. Not a medical device.
- No identity, face recognition, or continuous location tracking.
- Site tuning is required to manage alert volume.
Why it matters
- In nursing homes, pooled annual fall incidence ≈ 43% among residents. Source
- Unwitnessed falls drive long time-on-floor; in a ≥90 cohort, 30% lay >1 hour; all >1 h events were unwitnessed. Source
- Over 53% of older fall cases were still on the floor at ambulance arrival in a prospective cohort. Source
- EU estimate: ~1 in 10 fall-related injuries occur in residential institutions. Source
Stats are population-level; facility rates vary.
- Radar-based fall detection surveys report high accuracy in controlled settings; generalization to real-world remains a challenge. Source
- Recent mmWave studies show promising detection on lab datasets; results depend on environment and activity mix. Source
- Conclusion: privacy-preserving sensors are viable for alerting, but require site tuning to keep noise low. Source
These are modality-level findings, not NexaSense-specific claims.
FAQ
Does it record residents?
No video or audio. Signals are processed to generate alerts. Data retention is configurable.
What’s required to deploy?
Power, a mounting point, and basic network coverage or a local gateway. Typical install is ~30 minutes per room.
Speak with our team
Brief your requirements. We’ll validate fit and deployment timelines.
By contacting us, you agree to be contacted about NexaSense.
Not for diagnosis or prevention. Not a medical device.