RuView: WiFi Signals as Real-Time Spatial Intelligence

TL;DR: RuView reads channel state information from commodity WiFi hardware and turns it into spatial intelligence: room-level presence detection, breathing and heart-rate monitoring, and rough pose estimation — all without a camera. The firmware runs on ESP32 nodes and the processing pipeline is Rust.

Source and Accuracy Notes

• Repository: ruvnet/RuView
• License: MIT
• Stars: 93,867 (as of September 2026)
• Language: Rust
• Topics: awesome, claude, densepose, esp32, firmware
• Homepage: Cognitum.One/RuView

What It Is

RuView is an open-source stack for WiFi-sensing applications. Instead of relying on cameras or wearable sensors, it reads the channel state information (CSI) — the fine-grained signal data that WiFi chips expose for diagnostics — and runs signal-processing and machine-learning models on top of it. The result is a system that can answer questions like “is someone in this room,” “is that person breathing normally,” or “what rough pose are they in” using only the WiFi signals already flying around the space.

The project includes three layers:

• ESP32 firmware that captures CSI from compatible WiFi chips and streams it onward.
• A Rust processing pipeline that turns the raw signal stream into structured events: presence, motion intensity, breathing rate, heart rate, and dense pose coordinates.
• An integration layer that exposes these events via standard interfaces so home-automation, security, or health-monitoring apps can subscribe.

Why It Matters

Camera-based sensing is the default in smart homes, elder care, and security — and it’s also the source of most privacy objections. WiFi sensing sidesteps the camera entirely. The signal is already in the air; you just need the right hardware and software to read it.

A few reasons this category is interesting now:

• Hardware is cheap. An ESP32 node costs a few dollars. A four-node setup covers a normal room.
• It works through walls. WiFi signals penetrate drywall, so a sensing grid can cover multiple rooms from one access point.
• It runs at the edge. All processing happens locally; nothing has to leave the building to a cloud model.
• The privacy posture is genuinely different. No images, no audio, no light dependency. A signal that looks like “human present, breathing 14 breaths/minute” doesn’t carry visual identity.

The MIT license and the choice to publish the Rust pipeline openly make RuView a real foundation rather than a press-release demo. You can fork the firmware, port it to a different WiFi chipset, or hook the event stream into your own home-assistant config.

Practical Trade-Offs

WiFi sensing isn’t magic. A few honest limitations:

• Dense pose is rough. It’s not skeletal tracking in the way a Kinect or Apple Vision Pro gives you. You get a heat-map-style estimate of where bodies are, not joint-accurate motion capture.
• Vital signs need stillness. Breathing and heart-rate estimates work best when the subject is roughly stationary. Movement and pets will confuse the signal.
• Chipset support is uneven. Not every WiFi chip exposes CSI in a useful form. ESP32 with the right firmware is the well-trodden path; other platforms are less mature.
• Calibration matters. Multi-path effects from furniture and walls need to be characterized before the system gives reliable readings.

These are not deal-breakers for the use cases RuView is actually targeting — presence detection in a smart home, sleep monitoring, fall-adjacent elder care — but they shape what the system is and isn’t good for.

Who Should Care

RuView is worth a close look if you’re building:

• Smart home automations that need occupancy without cameras (lights, HVAC, security).
• Health monitoring tools that benefit from passive, always-on vital signs (sleep labs, elder care).
• Privacy-sensitive presence detection for offices, rentals, or shared spaces.
• Research prototypes in RF sensing, ubiquitous computing, or human-computer interaction.

It’s also a useful reference implementation for anyone learning about CSI, signal processing, or Rust pipelines for embedded-to-edge workloads.

Summary

RuView is one of the more substantive open-source efforts in the WiFi-sensing category. It pairs cheap ESP32 hardware with a Rust processing stack and exposes its outputs through standard interfaces. The privacy story is real, the license is permissive, and the use cases (presence, vitals, rough pose) line up well with the technology’s actual strengths.