WiFi Routers Can Now Monitor Your Breathing Through Walls
An open-source project called RuView turns ordinary WiFi signals into a real-time human sensing platform — detecting body pose, vital signs, and movement through walls, using hardware most people already own.
A standard WiFi router does more than push packets. Every time it communicates with a device, it collects metadata about how its signal traveled through the room — bouncing off walls, furniture, and people. An open-source project now exploits that metadata to reconstruct detailed human body poses, monitor breathing and heart rate, and detect presence through walls, all without a camera or wearable sensor in sight.
The project is called RuView, built by developer Reuven Cohen and hosted on GitHub. It implements a technique called WiFi DensePose, originally pioneered by Carnegie Mellon University researchers, as a practical edge computing system that runs on commodity hardware. Cyber Security News reported in March that the system "is turning ordinary WiFi infrastructure into a through-wall human-sensing platform detecting body pose, vital signs, and movement patterns without a single camera, raising urgent security and surveillance concerns."
That last part — the surveillance concerns — is where this gets complicated. The same technology that could monitor an elderly parent's breathing overnight could also track someone's movements without their knowledge or consent. WiFi sensing sits at a genuinely uncomfortable intersection of utility and intrusion, and its arrival as production-ready open-source code forces a conversation that's been mostly theoretical until now.
How WiFi Sees You
The technical foundation is something called Channel State Information, or CSI. WiFi hardware already collects CSI to optimize signal transmission — it's a detailed record of how radio waves propagate between a transmitter and receiver across dozens of frequency subcarriers. When a human body moves within that wireless environment, it distorts signal paths in measurable ways.
RuView's signal processing pipeline captures these disturbances at 54,000 frames per second using Rust, then extracts amplitude and phase variations from the CSI data. Those variations feed into a modified DensePose-RCNN deep learning architecture — a neural network model borrowed from computer vision — that maps the disturbances onto a model of the human body.
The output: real-time reconstruction of 24 body surface regions, including arms, torso, head, and joints, mapped to UV coordinates that mirror what a camera would produce. Except there is no camera. The system derives everything from radio frequency signals.
Vital sign monitoring runs as a parallel process. Bandpass filtering isolates specific frequency ranges: 0.1 to 0.5 Hz captures breathing patterns (roughly 6 to 30 breaths per minute), while 0.8 to 2.0 Hz filtering targets heart rate. The system essentially reads the tiny chest movements associated with respiration and cardiac activity from the way they perturb WiFi signals.
What Makes This Different
Researchers have been publishing papers on WiFi-based sensing for years. Carnegie Mellon's original WiFi DensePose work demonstrated the concept in an academic setting. What RuView does is package this into deployable, open-source software that runs on edge hardware — no specialized lab equipment, no proprietary sensors, no cloud dependency.
AIToolly noted when the project surfaced on GitHub's trending page in early March that the system "operates entirely without the need for video input, offering a unique approach to sensing human activity and health through walls using existing WiFi infrastructure." The emphasis on existing infrastructure matters. Most homes and offices already have WiFi routers. The marginal cost of adding this sensing capability is essentially the compute to run the model.
That accessibility is both the promise and the problem. Academic WiFi sensing research was gated by expertise and equipment. An open-source GitHub repo with documentation is not.
Healthcare Without Wearables
The most straightforward beneficial application is continuous health monitoring. Elderly patients living alone, post-surgical recovery at home, sleep apnea screening — all of these currently require either wearable devices (which people forget, remove, or find uncomfortable) or clinical visits (which are expensive and infrequent).
WiFi-based vital sign monitoring could run passively and continuously. A router in the bedroom could detect if someone's breathing pattern changes overnight, or if they've fallen and aren't moving. No watch to charge, no chest strap to wear, no camera recording your bedroom.
For healthcare systems already strained by aging populations and staffing shortages, ambient monitoring that requires zero patient compliance is genuinely valuable. The question is accuracy. Clinical-grade vital sign monitoring demands precision that WiFi sensing hasn't yet proven at scale outside controlled environments — the gap between compelling demo and FDA-clearable medical device remains wide.
The Security Problem Nobody Asked For
Cyber Security News framed the core tension bluntly: "Researchers and developers have long theorized that ambient radio signals could be weaponized for passive surveillance. That theory is now production-ready code."
Consider what this technology enables. Someone with access to WiFi signals — which propagate through walls by design — can potentially detect whether a building is occupied, how many people are inside, what postures they're in, and whether they're awake or asleep. No physical intrusion required, no visible surveillance equipment, and the sensing is entirely passive from the target's perspective.
For law enforcement and intelligence agencies, this is a capability that previously required thermal imaging equipment or other specialized tools. For stalkers and burglars, it's a new attack surface. For anyone living in an apartment with shared WiFi infrastructure, it raises questions about what your neighbors' router might be able to detect about your movements.
The open-source nature of RuView means the code is available for security researchers to audit and for defenders to understand the threat model. But it also means anyone can download and deploy it. This is the familiar dual-use dilemma of security research, accelerated by the fact that the sensing medium — WiFi — is everywhere.
Smart Homes Get Smarter, and Creepier
Beyond healthcare and security, WiFi sensing has obvious applications in smart home automation. Current systems rely on motion sensors (which detect presence but not identity or posture), cameras, or voice assistants (which require active interaction).
WiFi-based sensing could enable genuinely context-aware environments. A home that knows you're lying on the couch versus sitting at your desk versus standing in the kitchen — and adjusts lighting, temperature, and audio accordingly — without any visible sensors. Gesture recognition without cameras. Occupancy detection for energy management that knows the difference between an empty room and a sleeping person.
The smart home industry has spent years trying to make ambient computing feel natural rather than surveilled. WiFi sensing could advance that goal or undermine it entirely, depending on how it's implemented and governed.
What Comes Next
The immediate trajectory is predictable: more developers will build on this, hardware vendors will explore integrating CSI-based sensing into consumer routers, and regulators will scramble to figure out where WiFi sensing fits within existing privacy frameworks. Current wiretapping and surveillance laws were written for microphones and cameras, not for radio frequency body mapping.
Several things will determine whether this technology becomes a net positive. Accuracy validation against clinical standards will matter for healthcare adoption. Consent and disclosure frameworks will matter for everything else. The ability to detect and block unauthorized WiFi sensing — a defensive capability that barely exists today — will become increasingly important.
RuView, is a proof of concept that something previously confined to research papers now runs on edge hardware. The underlying physics aren't new. The accessibility is. And accessibility, in technology, is what turns curiosities into industries — and sometimes into threats.
The WiFi signals passing through your walls right now carry information about your body. The only question is who gets to read it.
