Researchers have developed mobile robots that can use Wi-Fi signals to effectively “see through” walls. It’s raising the possibility of flying drones using the technology to see inside buildings.
Led by Yasamin Mostofi, a professor of electrical and computer engineering, the group at the University of California, Santa Barbara (UCSB) has shown how the radio signals sent and received by a pair of wheeled robots can provide information about what lies behind concrete walls even when the objects do not move.
Depending on their properties, objects behind a wall will attenuate the radio signals. As the robots make multiple passes around on either side of a walled-off area, one robot measures signals broadcast by the other.
Differences in signal strength can reflect the presence of hidden objects, according to the researchers, who formulated a wave-propagation model for the project.
The approach, which has a targeted resolution of 2 centimeters, can show the location and shape of hidden objects, and has the potential to classify them as wood, metal or human.
Researchers at MIT previously developed a similar X-ray-style technology using Wi-Fi signals that they called Wi-Vi, but it’s designed to track moving objects or people. Mostofi’s group wants to be able to get a view of what lies behind thick walls made of brick or concrete.
“We have been interested in truly seeing through walls, meaning seeing every square inch on the other side with high accuracy, figuring out where all the objects are, as well as their geometry and material type, without any prior knowledge,” Mostofi wrote in an email.
“With Wi-Fi signals everywhere, it is very important to understand what they can naturally tell us about our environment. Also, with robots/drones becoming more and more part of our near-future society, it is just natural to see what the automation of this process can do.”
Mostofi’s group is now experimenting with a commercial aerial drone to see how well it can perform the X-ray feat. Multiple drones could be used to scan a concealed area, Mostofi said, or the receiver could be a ground-based robot or fixed node.
“In principle, this should also give us a nice 3D view, because you can change the height, as opposed to 2D imaging,” she said.
The UCSB system could be used for applications including search and rescue, detecting potential home intruders before entering a house or as health-monitoring systems for elderly patients.
One limitation of the technology is that as the area to be scanned gets more cluttered or larger, the accuracy of the imaging can go down, Mostofi said.
She and collaborators plan to work on improving the approach before it can be commercialized, which would depend on the application and the degree of accuracy needed.