RuView is a WiFi-based spatial intelligence platform written in Rust that transforms ordinary radio signals into real-time sensing data without requiring cameras, wearables, or cloud connectivity. The system detects human presence through walls, measures vital signs like breathing and heart rate, recognizes activities such as walking or falling, and monitors room occupancy by analyzing Channel State Information (CSI) captured from low-cost ESP32 sensors priced around nine dollars per node.
The core technology leverages the physics of radio wave disturbance. When people move, breathe, or remain stationary, they create measurable perturbations in WiFi signals that RuView's edge-deployed models interpret into actionable intelligence. The platform ships with 21 entities per sensor node, including eleven raw signal measurements and ten inferred semantic states such as someone-sleeping, possible-distress, room-active, elderly-inactivity-anomaly, meeting-in-progress, bathroom-occupied, fall-risk-elevated, bed-exit, no-movement, and multi-room-transition. Three starter Home Assistant blueprints are included for immediate integration.
RuView integrates natively with four major smart-home ecosystems. Home Assistant receives data via an HA-DISCO MQTT publisher, Apple Home and HomePod connect through a discoverable HAP-1.1 bridge, and Google Home, Amazon Alexa, and SmartThings access the system either through the Home Assistant bridge or via a Matter endpoint. Voice assistants can announce presence and vital signs by room without requiring custom skills.
The system runs entirely on edge hardware with no internet requirement. An ESP32 mesh paired with a Cognitum Seed provides persistent memory, cryptographic attestation via Ed25519 witness chains, and AI integration. The pretrained model, published on Hugging Face at ruvnet/wifi-densepose-pretrained, achieves 82.3 percent held-out temporal-triplet accuracy and fits in just eight kilobytes when quantized to four-bit precision, enabling microsecond inference on Raspberry Pi hardware. The platform includes a 105-cog catalog of edge modules covering health, security, building, retail, industrial, research, AI, swarm, signal, network, and developer applications.
According to GitGenius tracking data, the repository has grown from 76,426 to 76,428 stargazers since July 4, 2026. The project maintains active development with a median issue and pull request response latency of 1.2 hours across 361 tracked items, though the mean response time extends to 36.7 hours. The most active issue labels are bug with 65 items, enhancement with 56 items, and question with 35 items. Primary contributor ruvnet has logged 1,360 tracked events, with secondary contributors hcwang and munozluis1015 each recording eight events. The repository shares overlapping contributors with ruvnet/ruflo, ruvnet/ruvector, and nuxt/nuxt.
RuView's technical capabilities include real-time breathing rate detection via bandpass filtering at 0.1 to 0.5 hertz, heart rate measurement at 0.8 to 2.0 hertz, presence detection with sub-millisecond latency and thirty-second ambient calibration, seventeen-keypoint pose estimation achieving 82.69 percent torso-PCK@20 on the MM-Fi dataset, fall detection within 200 milliseconds, and through-wall sensing up to approximately five meters depending on signal strength. The platform supports multi-frequency mesh scanning across six WiFi channels and includes optional world model prediction using OccWorld TransVQVAE for fifteen-frame future occupancy forecasting.
