RuView
Open-source WiFi sensing platform that uses CSI from compatible ESP32 sensors to estimate presence, movement, breathing and heart-rate signals without cameras or wearables.
RuView turns changes in WiFi radio signals into a local sensing view. It uses Channel State Information (CSI) from compatible ESP32 sensors or research network cards to estimate presence, occupancy, movement, breathing and heart-rate signals without cameras or wearables. The repository also describes camera-free pose estimation, room fingerprinting, sleep monitoring, fall detection and multi-person counting. The project includes a sensing server, a Rust stack, Python training and evaluation tools, a browser-based Observatory, a dual-modal pose-fusion demo and live 3D point-cloud demos. It can connect to Home Assistant through MQTT and can be paired with Apple Home, Google Home, Amazon Alexa and SmartThings through the documented bridge and Matter integrations. The README lists 21 entities per node and three starter Home Assistant blueprints. To try live sensing, use CSI-capable hardware such as an ESP32-S3 mesh or a supported research NIC. The Docker image runs with simulated data for evaluation; an ordinary Windows, macOS or Linux laptop can provide only RSSI-based coarse presence and motion detection. The repository links to pretrained CSI weights and a live Observatory demo. RuView is beta software. The README says that the original ESP32 and ESP32-C3 are unsupported, a single ESP32 has limited spatial resolution, and camera-free pose accuracy is still limited. Vital-sign and pose outputs are research and prototype results, not medical devices, emergency systems or safety-certified controls; they require independent validation on the exact hardware, room, subjects and failure conditions. The project is released under the MIT license.