RuView is a Rust-based WiFi sensing platform that uses ESP32 edge hardware to detect presence, breathing, heart rate, movement, and room-level activity without cameras or wearables.
RuView is a spatial-intelligence and sensing system built on ordinary WiFi signals. According to the repository, it can infer presence, occupancy, breathing and heart rate, movement, room mapping, sleep-related signals, and other semantic states from radio disturbances, and it is designed to work with major smart-home ecosystems such as Home Assistant, Apple Home, Google Home, Alexa, and Matter-based bridges.
The project addresses the need for contactless, privacy-preserving room sensing that works through walls and in darkness, without cameras, wearables, or phone apps. It is positioned for home monitoring, occupancy detection, vital-sign tracking, and automation use cases where users want environmental awareness rather than visual surveillance.
Conceptually, RuView listens to how WiFi radio waves change when people or objects move through a space. It uses CSI captured from low-cost ESP32 sensors, then applies learned models and signal analysis to convert those disturbances into higher-level outputs such as presence, breathing, heart rate, pose, motion, and room state. The README also says the system can learn each environment locally, runs at the edge, and supports smart-home exposure through Home Assistant, Apple Home, Google Home, Alexa, and Matter-compatible pathways.
It is drawing attention because it combines a striking headline use case—seeing through walls with WiFi—with unusually broad smart-home compatibility and a strong emphasis on edge deployment. The repository also highlights a large ecosystem story around pretrained models, ESP32-based hardware, Matter/Home Assistant integration, and a fast-growing project profile, which makes it appealing both as a technical demo and as a practical home-automation platform.
The README itself points to related approaches rather than named direct competitors: CSI-based sensing, WiFi fingerprinting for room mapping, and learned WiFi pose-estimation models. It also references its own companion projects and artifacts such as RuVector, Cognitum Seed, a pretrained WiFi DensePose model, and a pose-estimation Cog, so those are the clearest comparison points visible in the repository; beyond that, the README does not provide a formal alternatives list.
AI-explained · grounded in each repo's README