© Paul Doyle / Alamy/Alamy
Unfortunately, most of these systems have huge drawbacks that would make anyone reticent to use them. GPS systems, for instance, can really only tell you a person's location, not how well they're doing. And cameras infringe on a person's privacy in a big way. That's why researchers have turned to technology that you're much more likely to see in an airport than in your aging parent's home: radar.
Radar, scientists say, is ideal for those who wish to remain independent in their own homes, but still like access to the security that comes with knowing someone or something is checking up on them. "The algorithm is smart enough to discriminate between two closely resembling activities - sitting and falling," says Moeness Amin, an engineering professor at Villanova University in Pennsylvania who is working on developing the system. "And once the radar declares that a person has fallen, it sends a signal to the mobile phones of family members and neighbors, as well as to first responders."
Designed primarily for people living alone, the system makes use of radar units that can fit in a hand. These units send out electromagnetic waves into the surrounding environment that return with modified frequencies when they hit an object or a person. "Each limb in your body, your head and your torso, reflects a frequency that is different from that one that was sent," Amin says. "And those changes are the markers that we can use to classify the type of motion." Amin and his team of researchers can therefore tell the difference between a person who is sitting down and someone who has fallen down with a cane in hand. They can also use the technology to monitor a person's heart rate and breathing. Moreover, the use of radar technology means that whoever receives an alert will be able to tell where a fall or a cardiovascular event occurred in a person's home.
Training the system
But the system, which is still in the development phase, is specific to the user, so it won't be easy to set up. Different people have different habits and gaits, Amin explains, so the algorithm needs to be trained to recognized the way someone moves around. If an elderly person uses a walker, for instance, the system needs to be able to recognize the metal contraption in front of that person's legs. This learning phase is extremely important because you wouldn't want the radar to mistake the "90-pound lab on the living room floor" for the person it's supposed to monitor, explains Kelly Nestor, an adult nurse practitioner and clinical instructor also at Villanova University.
Nestor has been sharing her expertise with the researchers to ensure that the system is capable of understanding how the elderly interact with their environment. This means that she has spent a lot of time in the lab teaching college students how to walk and fall like the elderly. Right now, she says close to 50 percent of elderly Americans are choosing to remain in their own homes instead of moving to a nursing home as they age, "so if we can do something to boost their safety, we will be in a much better place."
Unfortunately, training the radar to recognize how a certain person moves means that it won't work as well if they're doing something outside of their normal routine, such as entertaining guests. "We will train the radar for how Joe walks, how Joe sits and how Joe falls," Amin says, "so it will be very robust for Joe. But if I go and fall, then the false alarm rate won't be as small as it will be for Joe." And false alarms, Amin says, are a big concern. "Right now, we can get to one false alarm in 100, but that's using three units in the same room." To counter imprecisions, Amin suggests that people turn off the radar systems when they entertain guests. And if someone lives with a partner, that household will need more radar units to resolve both of their movements.
An uphill battle
Youngwook Kim, a computer engineer at California State University at Fresno who did not participate in the project, told The Verge in an email that "this radar system is valuable in that it can be an alternative solution when a camera system does not work in the dark to detect the falls." But Kim also said that he believes the system "should be improved" before it hits the market because it will require "a huge number of training databases" in order to control the rate of false alarms. In addition, Kim says that although "the idea itself is interesting, personally I am doubtful of how many elderly people will want to use this system" - especially since it needs to be installed "all over the place" in someone's home.
As Kim pointed out, the need for multiple radar units won't make it an easy sell, because having to outfit a home with multiple units means that the service will be extremely expensive. Amin says that he expects to see each unit priced at "less than $500," which means that covering an average-sized room will cost around $1,500. And the cost of the units themselves doesn't include the cost of the monitoring system that powers the units. So, in all likelihood, the service will resemble that of a house alarm. People will sign up and receive radar units at "no charge," Nestor says, but the customer will get billed for the service every month.
But the aspect that is most likely to halt the project's progress is public perception. "The safety of the electromagnetic waves is not a concern," Amin says, "but the psychological taboo of having a radar looking at you all the time will need to be worked out." Nestor believes that the alternative - moving to a nursing home where they lose most, if not all, of their independence - will be enough to make people consider using the radar system. "The number of Americans that are aging and who live alone, and own their own homes just continues to increase," she says, and "a move is traumatic when you are that much older."
Furthermore, there's always a chance that someone might use the system to monitor someone that doesn't want to be watched. After all, government and security agencies "have been doing that for the past 10 years," Amin jokes, so it can definitely be "misused and abused." But the researcher, who has represented academia at numerous NATO conferences, also thinks that if someone really wants to use radar to monitor another human being, they can already do so using the technologies that are readily available to the public. The system the researchers are developing, he says, is intended for very specific purposes. "We are monitoring the elderly and specifically focusing on falls and on anomalies of cross-motor activities."
A very real problem
Liu Liang, a computer engineer at the University of Missouri who did not work on the project but has done work on radar monitoring systems for the elderly in the past, thinks Amin's system could truly help people because falls "really cause fatal accidents" among the elderly. According to a study of Canadians over 65, about 62 percent of injury-related hospitalizations at that age are the result of a fall. Moreover, 1 in 1,000 falls results in death. So if a technology can save people's lives and protect people's privacy, Liang says, "I would recommend it to family and friends as we reach old age."
Amin and his team will perfect the system over the course of the next two and a half years, but they also plan on shopping around this fall for a company that can help them get the product to market. Unfortunately, even if it manages one day to improve the lives of the well-off retirees, it will be hard to tell because very few studies have actually investigated whether elderly monitoring systems have improved the quality of life of the people they monitor.
In fact, when I asked Nestor if she knew of any studies that have demonstrated a tangible benefit to the people who have used these devices, Nestor seemed rather surprised. "I don't know if there have been any qualitative studies that have been done about that at all, in terms of the elderly population," she says, adding jokingly, "I will tell my boss - we could start a new project."
Less personal relationships, more time to spend on facebook, texting on the smartphone...