It seems like there’s no limit to what we can do with WiFi these days, from syncing health data from a watch to our smartphones, to pressing a physical button to order more coffee. But now MIT researchers have found a way to use it to identify people, but not necessarily with the internet itself. Instead they’ve developed software that can detect people’s silhouettes through walls based on variations in WiFi signal.
The team at MIT’s Computer Science and Artificial Intelligence Lab built what they call RF-Capture, a technology that uses the reflections from its wireless signals to recognize a human form. While it’s been development for several years now, the researchers say it can now even detect the differences between body shapes and recognize individuals with up to 90% accuracy.
The RF-Capture device is used to emit WiFi signals, which can pass through a wall, and when they reflect off the body of someone standing on the other side, it captures images that are used to detect body parts via algorithm. Once enough images are collected to piece together a full silhouette, it can be used to analyze the small differences in shape and movement between individuals. The researchers say that in some cases they can event detect things as subtle as heart rate and breathing pattern.
While the technology is still in the early stages, and the MIT team hasn’t suggested any specific real-world uses at this point, it’s easy to see it being applied to various smart home features. For example, instead of pressing a button on a smartphone, or setting presets based on time, a home’s climate settings could automatically activated depending on which person gets home from work first.