Beta Intermediate

Aging-in-Place Guardian

Falls, bathroom emergencies and routine changes — sensed through walls, with total privacy.

What it takes

A validation-ready kit with a guided handoff.

Falls, bathroom emergencies and routine changes — sensed through walls, with total privacy. This kit deploys the LatentField 2.0 runtime for the Elderly world: an ESP32-S3 WiFi-CSI mesh feeds the same pipeline (compressed sensing, PCA + MVE anomaly manifold, radio-tomographic field, compartments) that was validated on 312 GB of public research data. It runs locally or in a container, pairs to your account with a claim-code PIN, and reports in plain language. Capabilities: presence, motion_level, breathing_rate, fall_detect, sleep_quality_proxy, zone_dwell, gait_signature. Explore them all at /v/elderly-care.

Starting hardware~$130

Typical setup60 min

Core hardwareESP32-S3 CSI mesh + LatentField 2.0 runtime

Data approachLocal by default

Choose hardware, setup, or support Design this space I already have hardware Watch it work

Explore freely. Create an account only when you want to save and deploy.

01Pair

One-time PIN adds the correct devices to this setup.

02Verify

Latent checks the live heartbeat before publishing.

03Hand off

The recipient receives a calm, branded app—not a control panel.

What is included

3× ESP32-S3 CSI node (Seeed XIAO / T-Dongle-S3)×3	~$27
Powered USB hub + USB-C cables	~$18
Runtime host — Raspberry Pi 5, mini-PC, or any always-on computer (or a container)	~$80
Estimated hardware total	~$130

Preview the installation

1Flash the CSI mesh nodes

2Run the LatentField 2.0 runtime

3Calibrate on an empty room

4Read the Elderly field

Open the step-by-step preview

Evidence, source and technical config

ReadinessBeta

Source projectlatent-field2.0

SensorsWiFi CSI

The full deployment config is exposed for technical review and can be copied or downloaded.

# LatentField 2.0 — Aging-in-Place Guardian runtime config # Runs on the host with: lf2 run --config configs/verticals/elderly_care.yaml pipeline: { vertical: elderly_care } capabilities: - presence - motion_level - breathing_rate - fall_detect - sleep_quality_proxy - zone_dwell - gait_signature zones: - { name: bedroom, shape: rect, x0: 0, y0: 0, x1: 3, y1: 4 } - { name: bathroom, shape: rect, x0: 3, y0: 0, x1: 5, y1: 2 } - { name: living, shape: rect, x0: 3, y0: 2, x1: 6, y1: 5 } manifold: { baseline: activity } # lived-in space: routine is normal, departures alert

Reviews and field notes (0)

No field notes yet. Evidence and readiness are still shown above so an empty review list never pretends to be proof.