One of the most daunting challenges for seniors living on their own could be dealt with by something as simple as a wireless network.
A new system developed by researchers at the University of Utah targets the leading cause of injury-related death among people 65 and older—falls.
As the population of the U.S. ages, with a large number of baby boomers already in their golden years, this will likely become even more of an issue.
“The costs of nursing home care are so high, and people generally want to live independently,” says study author Neal Patwari, Ph.D., an associate professor of electrical and computer engineering at the University of Utah. “It makes sense to deploy some inexpensive sensors that can detect falls and call for help if the person can't themselves.”
Using Wireless Has Advantages
The sensors chosen by Patwari and computer engineering graduate student Brad Mager are similar to those used in home Wi-Fi networks. The researchers deployed these small, low-cost radio-frequency (RF) sensors around a room at two different heights. This enabled them to determine whether a person in the room was falling.
Unlike other systems that detect falls, the RF sensor network doesn’t require people to wear a device, such as the familiar medical alert buttons used by many elderly people.
“Most people who own one of those emergency call buttons or fall sensors aren't wearing them at the time of their fall,” says Patwari, “so they are not particularly useful.”
In addition, the RF sensor network is more respectful of privacy because it can detect only the location of objects larger than six inches. Fall detection systems that rely on video surveillance, on the other hand, can make people feel uncomfortable because they take continuous and detailed recordings.
Building a 3D Image of a Person
The system works because the human body is largely made up of water. So as a person stands in a room, his or her body alters the strength and path of the wireless signals as they pass from one sensor to another.
The radio waves, however, are not blocked by non-metal walls or furniture, meaning the system can “see through” most obstructions. The sensors can also be hidden behind walls or inside other objects.
The information gathered from the wireless network can be used to pinpoint the location of a person in the room. In essence, the researchers are able to take many one-dimensional measurements between the sensors and convert them into a three-dimensional image of a person—a technique that’s called “radio tomography.”
Tracking Falls from a Rough Image
By placing the sensors at two levels, they are also able to determine if a person is standing, sitting, or lying on the floor.
“I thought that if we put sensors at different heights, it might be possible to locate someone with the same accuracy in three dimensions, and then potentially detect a fall,” says Patwari. “That is, we might have the potential, with our radio tomography technologies, to detect falls without requiring the person to wear anything.”
The rough image of the person appears as five separate layers. Falls are detected, then, by seeing how each layer changes over time. For example, a person falling would quickly disappear from the top layer, followed by taking up more space in the bottom layer.
The researchers conducted experiments to train the system to identify different movements within the space—falling, sitting down, and lying down on the floor.
Detecting More Than Just Falls
The system, which was presented Sept. 10 at a meeting of the Institute of Electrical and Electronics Engineers, is still in the early stages. Patwari hopes to develop it into a commercial product through his start-up company, Xandem Technology.
In addition to alerting caregivers or emergency services if an elderly person falls, the system could also be tied into a larger home sensing system that tracks the overall health of those living there.
“Our same RF sensors deployed for fall sensing can simultaneously be used for room-level tracking and breathing rate monitoring, as we have demonstrated in past work,” says Patwari.