Home, safe home
Home isn’t just where the heart is—it’s also where potential hazards lie. That’s especially true for the elderly. More than 40 million people in the United States are age 65 or older, a number expected to increase to 55 million by 2020. Moreover, about 31 percent choose to live alone.
To help keep seniors safe in their homes, researchers at the Heracleia Human-Centered Computing Laboratory in the Engineering Research Building explore ways to unobtrusively monitor their daily activities and alert caregivers when help might be needed.
Led by Professor Fillia Makedon, chair of the Computer Science and Engineering Department, the scientists have set up a mock apartment at Heracleia (named after a city in antiquity where Hellenic culture thrived) outfitted with sensors that transmit radio frequencies. These sensors—placed on beds, thresholds, TVs, the refrigerator, a “smart drawer” where medication is stored—are matched to a pattern of daily activities. If something unusual is detected, an alert goes out.
“Our ultimate goal is to facilitate ‘aging in place,’ where people can decide the level of monitoring support they wish to have to feel better about staying at home alone, even if it’s for just a few hours at a time,” Dr. Makedon says.
Sensors in an apartment can capture movement, temperature, sound, and other elements. They can be worn to help indicate if the elderly resident has fallen or isn’t responding. Additionally, the researchers are working on assistive robots to recognize the person’s position and follow him or her into places where no other sensors are available.
“The sensors look out for those moments—‘events,’ as we call them—when a big change or something unexpected occurs. If a big event takes place, such as a fire alarm sounding, then a central location that is in charge of wirelessly collecting signals from different elderly apartments would be alerted,” Makedon explains. “The event could also be an absence of something, such as a person not going into the kitchen at all, which would indicate that he or she has not eaten.”
The researchers’ work could enable a son, daughter, physician, or other caregiver to monitor from afar whether an elderly person got out of bed, took medication, remained active throughout the day, watched TV, ate, went to bed, and slept restfully.
Privacy is a key issue—ensuring the data is secure and deciding who should see it. “One way we’re dealing with that is by not using cameras unless specifically requested by the patient or doctor,” Makedon says. “The sensors we use give out data that is seemingly not important unless one has the capability to put it all together to make sense of the readings.” Plus, the sensors can be deactivated when their presence isn’t required—for example, when people come to visit.
While Makedon and her team aren’t yet testing the sensors on live subjects, they are collaborating with University of North Texas Health Science Center experts who work with nursing homes. And the project is already yielding promising, and unexpected, results.
“We are currently using pressure pads to detect unusual sleep patterns, which doctors think may indicate a higher risk for ‘wandering behavior,’ ” Makedon says. “But we have found that this setup can also capture people’s seizures. So our technology is turning out to have multiple beneficial uses, which is very encouraging.”