One block per resident with vital trends (pulse, blood pressure, SpO₂, respiration), location and events. Includes a simulated emergency sequence: fall → stroke-typical vital pattern → cardiac arrest.
Four core elements per patient — not 50 numbers, but a clear picture with trend.
Pulse, blood pressure, SpO₂ and respiration as a mini-trend per resident. Caregivers see at a glance whether a value is "tipping", not only whether it is currently too high.
CSI-based (bed, armchair, by window, on the floor, room empty). No camera, no microphone — only the reflection of existing WiFi signals.
Falls, restlessness episodes, unusual standing-up patterns, vital decompensations are marked as hints and reconstructed over time.
Multi-stage: ward display → push to on-duty staff → optional emergency call. Which stage triggers when is defined by the facility — no black-box automation.
Short answer: experimental only, not for daily care.
Tissue water content affects dielectric permittivity — and therefore radio propagation in theory. Early research (mmWave / UWB / WiFi-CSI) shows correlations between CSI patterns and hydration state under controlled lab conditions: probands before/after drinking water, well-defined poses, stable environment.
In real care settings — multi-bed rooms, moving caregivers, changing furniture, other WiFi devices — the signal is so noisy that a reliable hydration statement is not possible today. We promise what the technology can do today; hydration belongs on the research roadmap, not in the MVP.
Details and sources on the WiFi-CSI page.
Falls · stillness · notable vital patterns · presence. That is reliable today.
Stroke, cardiac arrest, hydration, pose estimation & friends: recognisable patterns, but not a diagnosis. The caregiver decides — we just give them the right information in time.