The Wi-Vi system relies on two antennas to broadcast Wi-Fi signals and a receiver to read them, according to the researchers’ paper. The Wi-Fi signals degrade in quality each time they pass through a wall, so the receiver must be prepared to pick up on very weak signals. It is also quickly overwhelmed if there are too many to sort through.
To combat this, researchers at MIT developed a two-signal approach. The signals are inverse, so when they travel the same path they cancel each other out. When they hit a moving object, though, such as a person behind a wall, they are reflected at different angles. The receiver picks up just these changed signals, as the rest canceled each other out.
A person’s movement is tracked based on changes in distance from the receiver. The longer it takes for the reflected signal to make it back to the receiver, the farther away the person is. The system then generates a graph of their position over time. It is not a literal image of the person, but it can be used to infer if they are walking, running or something else. Zero movement registers as a flat line, but minor movements are easily picked up.
Each bump in this graph represents a step. The bumps above the red line indicate a step forward, while the ones below the red line indicate a step backward. MIT
Instead of tracking another person, the Wi-Vi system can also track the user. For example, you could wave your arm to turn off the lights in another room. The system was able to differentiate between two separate hand gestures 100 percent of the time through glass, a hollow wall and a wood door during testing.
Researchers think the Wi-Vi system could also be used to find survivors in destroyed buildings or count and track criminals. Compared to previous military-oriented tracking systems, Wi-Vi is cheap, compact and lightweight, which makes it practical for consumer uses such as personal safety. Researchers now hope to improve the technology so it can work with denser walls over longer ranges.
Wi-Vi isn’t the first system to use everyday Wi-Fi to detect motion. The University of Washington’s WiSee interface detects gestures to control connected devices in the home.