University of Utah engineers have developed a system that uses a wireless network of radio transmitters to track people moving behind solid walls. They say the system could help police, firefighters and other emergency services capture intruders, and rescue hostages, fire victims or elderly people who fall in their homes by letting them know where to focus their attentions. The engineers' system uses radio tomographic imaging (RTI) to "see", locate and track people or objects in an area surrounded by inexpensive radio transceivers that send and receive signals.
Unlike radar, which reads the reflections of radio signals bounced off targets, RTI measures the "shadows" in radio waves created when they pass through a moving person or object. By measuring the radio signal strengths on numerous paths as the radio waves pass through a person or object a computer image can be constructed.
One of the studies carried out by the researchers to test the system involved placing a wireless network of 28 inexpensive radio transceivers - called nodes - around a square-shaped portion of an indoor atrium and a similar part of a grassy area with trees. In the atrium, each side of the square was almost 14ft long and had eight nodes spaced 2ft apart. On the lawn, the square was about 21ft on each side and nodes were 3ft apart. The transceivers were placed on 4ft-tall stands made of plastic pipe so they would make measurements at human torso level.
Radio signal strengths between all nodes were measured - first when the rectangle was empty and then as a person walked in each area. They developed math formulas and used them in a computer program to convert weaker or "attenuated" signals - which occur when someone creates "shadows" by walking through the radio signals - into a blob-like, bird's-eye-view image of that person walking.
A second study detailed a test of an improved method that allows, "tracking through walls". The study details how variations in radio signal strength within a wireless network of 34 nodes allowed tracking of moving people behind a brick wall.
The method was tested around an addition to assistant professor, Neal Patwari's Salt Lake City home. Variations in radio waves were measured as fellow researcher, Joey Wilson walked around inside. The system successfully tracked Wilson's location to within 3ft.
The wireless system used in the team's experiments was not a Wi-Fi network like those that link home computers, printers and other devices. Rather, it was a network-based on Zigbee, a high level communication protocol for wireless personal area networks (WPANs) that can be found in home automation devices like the Control4 Wireless Dimmer and remote control devices like the Roboni-I robot.
Since RTI uses radio frequency (RF) signals the system will work where optical and infrared imaging systems cannot, such as in the dark and through obstructions such as walls, trees and smoke. Also the researchers say their RTI system reduces privacy concerns because the current images produced cannot be used to identify a person.
In dangerous scenarios, like a hostage situation or fire, Patwari envisages that, before attempting to enter the building, police and firefighters would first place dozens of the radio transceivers around the building and would be able to see a computer image of where people were moving inside. Although Patwari notes that, in the case of a burning building, the system would also "see" moving flames, he points out that this could help to track where a fire was spreading.
Aside from the emergency applications the system could also be used by marketers to indicate where shoppers spend their time in stores and ascertain whether a marketing display is in the most effective location.
Border security in dark and foggy areas and home automation applications such as controlling lighting, heating and air conditioning or even controlling a sound system to direct the best sound to where people are located are other possible uses for the technology, say the researchers.
The University of Utah researchers RTI system is detailed in the study, Through-Wall Tracking Using Variance-Based Radio Tomography Networks.
The following videos demonstrate the technology both with and without walls.
