The same Wi-Fi technology that allows us to connect to the Internet without wires may soon let us see through walls or turn up the air conditioning with a wave of the hand.
Scientists at the Massachusetts Institute of Technology and the University of Washington are developing systems that detect Wi-Fi radio waves as they bounce off objects or people. Essentially they can turn a Wi-Fi device into a kind of radar, that can peer through walls and detect motion as far away as 30 feet.
The technology suggests that Wi-Fi devices could serve as a new kind of remote control, capable of responding to hand gestures. Or they could become lifesaving tools for firefighters, police, or soldiers who need to know if there’s a human being on the other side of a closed door.
“There was great interest in these kinds of technology from the army and military research,” said Dina Katabi, an MIT professor of electrical engineering and computer science. “But recently people got interested in whether you could do this for ordinary people.”
Wi-Fi is a low-powered radio system for connecting wireless devices to the Internet or to each other. Wi-Fi is used in billions of network routers, laptops, smartphones, video game consoles and other devices, so the necessary chips are cheap and easy to get.
Katabi and graduate student Fadel Adib call their detection system “Wi-Vi.” It works much like radar, by transmitting radio waves that bounce off various objects, including people. The trick is to build a receiver that can pick up the reflected waves, discard the unwanted signals, and identify those that are bouncing off humans.
For example, much of the radio energy that hits a wall simply bounces back, rather than penetrating into the neighboring room. This reflected energy is much more powerful than the radio signal bouncing off a human on the other side of the wall. Wi-Vi solves this problem by broadcasting two Wi-Fi signals, one out of sync with the other. When these signals bounce off stationary objects like walls, they cancel each other out. But bounces from moving objects, like people, come through clearly.
“You can remove the effect of the wall but you can see the object behind it,” Adib said.
The data are processed and fed to a video display, which indicates the presence of a moving person and his approximate location inside the building.
“At this stage, the technology doesn't allow you to see body parts or the face of the person,” Katabi said. “You just see a blob.”
Indeed, Katabi admits that today’s version of the system can’t tell the difference between a human being and a dog. She said that problem will be solved as image resolution gets better. Katabi said Wi-Vi probably won’t produce crystal-clear images of people, but it may well become sharp enough to pick out facial features such as noses and ears.
Meanwhile, one of Katabi’s former students, Shyam Gollakota, an assistant professor of computer science at the University of Washington, is developing WiSee, a Wi-Fi-based detector that consumers may someday use to control various household devices using gestures.
Existing gesture-based systems like Microsoft Corp.’s Kinect video gaming device aim cameras at the user to track his body movements, and so need a clear line of sight to work. WiSee can work even when the user is out of camera range or in a different room. WiSee could let a user switch on a TV with a wave of the hand or dim the lights by twirling a finger. But Gollakota said it could also be a valuable new home safety tool for elderly or disabled people.
“Imagine that your wireless router figures out you fell down,” Gollakota said. “You could automatically call 911.”
Gollakota said that he’s received inquiries from several electronics companies hoping to build WiSee technology into their products. “We were surprised by the kind of attention this whole project got,” he said.
This next wave of Wi-Fi technology does raise privacy issues, as such devices might be used by police to track people’s activities inside their homes without permission. Katabi said that Wi-Vi devices might contain a feature that would send a warning to people in the neighborhood that they were being watched.