Your WiFi router is constantly sending invisible signals through the walls of your building. Normally those signals just carry internet traffic. But a project called WiFi DensePose figured out how to read the tiny disturbances those signals make when a human body moves through them — and turn that into a surprisingly detailed picture: where someone is standing, how fast they're breathing, even their heartbeat.
No camera. No microphone. No footage of anyone.
The obvious use cases are genuinely useful: detecting if an elderly person has fallen, finding survivors after an earthquake, or simply knowing whether a room is occupied without installing sensors everywhere. But the quieter implication is that WiFi — something almost every building already has — is now a sensing layer that software can read.
This project has over 15,000 people watching it on GitHub right now, and the community around it is very active this week with new tutorials and ideas.
Not something to install yourself today. But if you run a care facility, a hotel, a retail space, or any building where knowing how people move through it would change how you operate — this is worth keeping on your radar. The technology is moving fast, and the people building real products on top of it are just getting started.
Open-source — software where the recipe is public and free. Anyone can read it, use it, or improve it.
CSI (Channel State Information) — the technical name for the detailed signal data your WiFi router is already collecting about its environment. Think of it as the router noticing the world around it.
Docker image — a ready-to-run package of software, like a lunchbox that already has everything inside. You don't build it yourself; you just open it.
GitHub stars — a rough measure of how many people in the tech world find something interesting or useful. 15,000 is a lot.