WiTrack system allows for motion tracking through walls
Microsoft's Kinect system is certainly impressive, but now that we've had a chance to get used to it and start taking it for granted, it does have one problem – you have to stay located in front of it. MIT's new WiTrack system, however, can track users' movements even when those people are in another room. Among other things, this could allow for video games in which the players run all over their house.
WiTrack incorporates one set of four antennas, which should be sufficient to cover all the rooms in one house. One of those antennas transmits radio signals (right through walls), which are subsequently reflected off the subject's body. The other three antennas receive those reflected signals.
Because the locations and distances between the three receivers are already known, the location of the subject can be triangulated by comparing how long it takes between the transmission of each signal, and its reflection being picked up by each of the receivers. Algorithms filter out distracting reflections from furniture and walls, leaving only the distinctive reflections made by a human body.
The system can reportedly track a person's location in real time to an accuracy of 10 to 20 cm (4 to 8 in), on x, y and z axes – in three dimensions, in other words. It can even track the position of parts of their body, and can recognize hand gestures.
The radio signals utilized by WiTrack are reportedly "100 times smaller than Wi-Fi and 1,000 times smaller than what your cell phone can transmit." A similar MIT-developed system known as WiVi does use Wi-Fi signals, but is actually less accurate because of Wi-Fi's more limited bandwidth.
Along with its potential for gaming, other suggested applications for WiTrack include location monitoring and fall detection for the elderly, and systems in which users could control appliances in other rooms via hand gestures.
WiTrack was developed by WiVi creators Prof. Dina Katabi and her graduate student Fadel Adib, working in collaboration with Prof. Rob Miller and his graduate student, Zach Kabelac. The system is demonstrated in the video below.