How ambient intelligence works
Continuous, dignified, clinical-grade monitoring, without a single camera.
What it is
Sensing without seeing
The environment itself observes. Not a person, not a camera, not a wearable. The sensor reads position and motion, never a face.
Why not cameras?
Identity
Camera footage is inherently identifiable. Ambient sensing data is not.
Consent
Residents with dementia cannot meaningfully consent to being watched on video.
Capacity
Cameras demand network and storage most LTC facilities cannot support.
Darkness
Our sensing works in the dark, and through bedding. Cameras do not.
Not her face, but the soft trace of every breath and movement, enough to know she is safe, and nothing more.
Architecture
The three-layer stack
01 · Hardware
The sensing layer
A contactless sensor in each room continuously senses presence, movement, and micro-motion, without cameras or microphones. It never forms a photographic image.
02 · Cloud
The inference layer
Sensor data is processed in the cloud, where proprietary signal pipelines extract clinical features: fall events, respiration rate, sleep staging, behavioral patterns. All computation is HIPAA-compliant and stays in a per-tenant environment.
03 · Application
The care layer
Ella AI translates inference into clinical language. Staff receive alerts, summaries, and trend reports through a simple dashboard and mobile notifications. No training required, no new workflow imposed.
Evidence
The science
Patent · International filing
2025
PCT/US2025/61807
Methods and systems for continuous non-contact physiological monitoring in long-term care environments.
Grant · Research funding
ODAT
University of Minnesota
An Office of Discovery and Translation grant supporting the translation of ambient-sensing research toward clinical deployment in long-term care.
Ready to see the clinical evidence?
How we are validating ambient sensing against clinical ground truth.