Date: Oct 15, 2009 Author: http://www.economist.com/node/14637299?story_id=14637299 Source: Economist Magazine
SUPERMAN had X-ray vision, which was useful for looking through walls when rescuing heroines and collaring villains. But beyond Hollywood, the best that engineers have been able to come up with to see inside buildings are devices that use radar. Some are portable enough to be placed against an outside wall by, say, a police unit planning a raid—and sophisticated enough to show, with reasonable accuracy, the location of anyone inside. But the best models cost more than $100,000, so they are not widely deployed. Now a team led by Neal Patwari and Joey Wilson of the University of Utah has come up with a way to peer through the walls of a building using a network of little radios that cost only a few dollars each.
Radar works by recording radio waves that have been reflected from the object under observation. Dr Patwari's and Mr Wilson's insight was to look not for reflections but for shadows. Their device broadcasts a radio signal through a building and, when that signal comes out the other side, monitors variations in its strength. The need for variation means the system cannot see things that are stationary. When the signal is temporarily blocked by a moving object such as a person, however, it shows up loud and clear.
Using a network of small transmitters and receivers, the researchers have found it is possible to plot a person's position quite accurately and display it on the screen of a laptop. They call the process radio tomographic imaging, because constructing an image by measuring the strengths of radio signals along several pathways is similar to the computerised tomographic body-scanning used by hospitals—though medical machines employ X-rays, not radio waves, to do the scanning.
The radios used by Dr Patwari and Mr Wilson are low-cost types designed for use in what are known as ZigBee networks. In that application they transmit data between devices such as thermostats, fire detectors and some automated factory equipment. They are not even as powerful as the radios used in Wi-Fi networks to link computers together.
Small and inexpensive as these ZigBee radios are, though, there is strength in their numbers. Each is in contact with all of the others. A building under examination is thus penetrated by a dense web of links. In one experiment, for example, a network of 34 radios was able to keep track of Mr Wilson's position with an accuracy of less than a metre—a figure that Dr Patwari and Mr Wilson think could be improved greatly by using specially designed radios instead of off-the-shelf ones. Moreover, putting radios on the roof of a building as well as around its walls should make it possible to produce three-dimensional views of what is going on inside.
The ability to “see” people moving around in a building with such a cheap system has many plausible applications, and Mr Wilson has set up a company called Xandem to commercialise the idea. Besides military, police and private-security uses, radio networks might be employed to locate people trapped by fire or earthquake. More commercially, they might be used to measure what retailers call “footfall”—recording how people use stores and shopping centres. At the moment, this is done with cameras, or by triangulating the position of signals given off by mobile phones that customers are carrying. Radio tomography could be simpler, more accurate and, some might feel, less intrusive. Certainly less so than a man in tights with X-ray eyes.
