ruvnet/RuView

What it solves

RuView transforms ordinary WiFi signals into a spatial intelligence system, allowing for the detection of people, movement, and vital signs without the need for cameras, wearables, or cloud connectivity. It enables non-invasive monitoring through walls and in the dark, solving the privacy concerns associated with traditional visual surveillance while providing critical data like breathing rates, heart rates, and fall detection.

How it works

The system uses Channel State Information (CSI) from low-cost ESP32 sensors to capture disturbances in radio waves caused by human presence and activity. These signals are processed using a combination of of spiking neural networks for local adaptation and pretrained models (available on Hugging Face) to infer semantic states. It can be deployed as an ESP32 mesh or paired with a Cognitum Seed for persistent memory and AI integration, integrating natively with smart-home ecosystems like Home Assistant, Apple Home, and Google Home.

Who it’s for

It is designed for smart-home developers, researchers in RF sensing, and those building healthcare or security applications that require contactless, privacy-preserving monitoring of occupancy, vitals, and activity.

Highlights

• Contactless Vitals: Real-time monitoring of breathing (6–30 BPM) and heart rate (40–120 BPM).
• Camera-Free Pose Estimation: Estimates 17 body keypoints from WiFi CSI signals.
• Privacy-First Edge AI: Runs entirely on edge hardware (ESP32/Raspberry Pi) with no cloud or internet required.
• Broad Integration: Native support for Matter, Home Assistant, Apple Home, Google Home, and Amazon Alexa.
• Extensive Module Catalog: Includes over 100 "Cogs" for specific use cases like fall detection, sleep quality monitoring, and multi-person counting.