Living smarter with sensor technology
WSU professors study ways to help seniors remain independentSeptember 22nd, 2016
For many people approaching retirement, finding ways they can continue to live at home safely as they age is an issue, one for which two Washington State University professors at the university’s campus in Pullman have been seeking a solution.
For almost 10 years now, professors Diane Cook and Maureen Schmitter-Edgecombe have been researching what’s called smart-home technology, a kind of sensor monitoring system that might help seniors stay independent longer. A study that’s part of that long-term effort currently is underway and is scheduled to wrap up next summer.
Cook, a professor of electrical engineering and computer science, says she first became interested in smart-home technology in 2007.
“My in-laws are in the real estate market and told me about homes with appliances that could talk to each other,” she says.
A smart appliance is any device that is able to connect to a smartphone, a tablet, or a computer, and pass along information to the user. A smart clothes washing machine, for example, could send out an alert via text message when a load of laundry is finished.
Cook’s work focused on artificial intelligence, machine learning, and smart environments, but she was interested in finding ways of using technology for health monitoring and intervention. She shared some of her ideas with Schmitter-Edgecombe, a professor of clinical neuropsychology, who had been studying memory intervention and the relationship between memory deficits and the capacity to perform complex daily tasks.
The two saw that a research partnership could combine their two fields of expertise and put them on the path to creating a smart-home for seniors.
“I had done clinical work with older adults, working to develop ways for them to stay independent, and studying how cognitive decline affects everyday living,” says Schmitter-Edgecombe. “Diane had a similar goal, in finding a way for technology to help keep people independent in their homes.”
Schmitter-Edgecombe and Cook wanted to start by working with people who had cognitive disabilities. To keep things simple, they also wanted the sensor technology to be part of the test environment, rather than a wearable device.
“It was hard to find ways to incorporate hardware into a home, because when we started, there weren’t really any wireless sensors,” says Cook. “Luckily, over time, wireless sensors have developed that are much lighter and simpler, easier to hide amongst a home environment.”
To begin testing sensor hardware that could monitor activities, as well as the software that could read the data, the professors turned a student apartment on WSU’s Pullman campus into their test lab.
“The lab is an actual apartment, with students who live there during the year,” says Cook. “While the students are out during the day, we conduct studies.”
Cook says the three-bedroom apartment is roughly 1,100 square feet, and is outfitted with sensors that use infrared technology to track motion and temperature, among other things. Data from the sensors is then collected and analyzed at WSU’s Center for Advanced Studies in Adaptive Systems, also called CASAS.
“We were primarily interested in tracking movement throughout the home. That—combined with the day, time, and patterns—is sufficient to understand what’s being done,” Cook says.
While there are no video or listening devices included, the sensors are able to track the motions involved in daily activities, such as bathing, sleeping, cooking, and eating, as well as social interactions, such as making a phone call.
“Once we put the hardware in place, we moved on to gathering data and recognizing activities from the readings given through the motion sensors,” she says. “We could then use that data to identify what people are doing and understand it. From there, we started to use it to access their health. Are they healthy and functioning independently? If not, the next step is thinking of ways to automate things, to control the home and help them.”
Schmitter-Edgecombe says the study began by using undergraduate students and trying to recognize activities based on their movements within the home.
“We then looked at more complex activities, and other factors like adding more than one person in the environment,” she says. “From there, we moved on to using older adults with mild cognitive impairments so we could see the quality of their performance of tasks, identify the difficulties, and begin looking at how to assist.”
The duo’s work with the pilot project set the stage for further funding through a $800,000 Washington state Life Sciences Discovery Fund in 2009.
Since that time, their research has received funding from individual companies, and grants from the National Science Foundation and the National Institutes of Health.
Following the pilot project, which concluded in 2009, Schmitter-Edgecombe says the two began to focus their research on how technology could be used to support individuals in completing tasks.
“We started to look at whether a verbal or visual prompt could help cue people, or remind them when they get stuck completing a task,” she says.
Further research into this area was funded by a $900,000 grant from the National Science Foundation to design an easy-to-install “smart home in a box.”
“The funding allowed us to design a kit that could be mailed to people who wanted to help us collect data,” says Cook.
The project has a website, casas.wsu.edu, where interested individuals can request to participate.
Cook says the kits take about an hour to set up and include 20 to 25 sensors that monitor and learn residents’ daily activities, note any changes, and prompt them if they forget to do something. The study will continue until next summer.
According to Cook, through that project and others, more than 100 smart home test sites have been set up in 11 countries to date.
“Some of the data is used to design computer science programs to better see patterns in data or look at differences in behavior across various demographics,” Cook says.
“Humans have really varied, complex behaviors. The more data we have, the more we can come to understand the impact those behaviors have on health.”
While a cost for the smart-home system is difficult to estimate due to constantly changing technology, Cook says there has already been significant interest among both researchers and private companies in how a smart home package designed for use by seniors and people with disabilities could be marketed to the public.
“It’s still pretty challenging to educate people on the program’s usefulness, but it is gaining in popularity,” she says. “We get visits from companies who want to talk about it at least once a month.”
Cook says that this smart home system differs from others already on the market in that it is more accurate, and is designed to actively learn about new tasks based on the user’s history.
In addition to their work on the smart-home-in-a-box project, this year the two professors also were awarded a $720,000 grant from the U.S. Department of Defense, targeted toward providing aid to people with cognitive impairment.
The grant enabled them to design a digital memory notebook, the first prototype of which was created this year.
Cook says the tool was based on Schmitter-Edgecomb’s years of work with paper-based memory notebooks.
“It works together with the smart home to inform individuals of the activities they already performed that day and to remind them of upcoming activities at the times they normally perform them,” says Cook. “This information is based on sensed activities so the users do not have to program the aid.”
Although there is already some technology out there that can help seniors to live more independently at home, Cook says it still isn’t as robust as she would like to see.
“I suppose as researchers, we’re never satisfied, we’re always looking at the next step,” she says.
Looking ahead, she says she and Schmitter-Edgecombe want to continue to pursue ways in which the smart-home technology can be used for preventive health care interventions.
“This has the potential to detect health events like falls and heart attacks, and inform caregivers,” she says. “It also could detect slower changes that may indicate cognitive decline, or mobility issues.”
Similarly, Schmitter-Edgecombe says, “For preventive care, prompting could be used to help people begin to integrate new healthy behaviors into their everyday lives.”
People interested in being part of smart-home technology studies can contact WSU’s Aging and Memory Laboratory at 509-335-4033.