What if slight changes in your facial expression, posture, or how often you eat are enough to determine whether you’re at risk for developing a disease or mental disorder? How would you spot these early-warning signs? One research group in Bristol, England, has designed a smart home that could help its inhabitants gauge such factors over time.
Known as the SPHERE (Sensor Platform for Healthcare in a Residential Environment) House, it’s an old, two-story Victorian home that the project’s researchers retrofitted with smart technologies to determine if data it collects on the occupants can track their health. The group designed the home with an integrated and synchronized system of some 60 sensors, cameras, and wearables. The project received funding of a little more than US $18 million from the Engineering and Physical Sciences Research Council, the United Kingdom’s equivalent to the U.S. National Science Foundation. The goal is to have the system installed in 100 homes in the Bristol area by 2017.
“This system, which is invisibly embedded in the house, monitors changes that might be correlated with certain health conditions,” explains IEEE Fellow Ian Craddock, who is leading a research team composed of several IEEE members. Craddock, an engineering professor at the University of Bristol, is also managing director of the Toshiba Telecommunications Research Laboratory, in Bristol.
“Many long-term health conditions are highly correlated to lifestyle,” Craddock says. For example, if someone is sleeping more than they used to, eating less, and slouching, it may be a sign they are developing depression. Eventually, the system could point out key behaviors that have changed over time.
The 60 researchers on the team come from universities in Bristol, Reading, and Southampton, as well as from IBM, Toshiba, and Bristol City Council. The two-bedroom house that the team retrofitted is owned by the University of Bristol. Researchers have been occupying the house for the past few months to ensure the sensors and cameras are tracking occupants accurately. They are now recruiting volunteers, including families with children, to move in for several weeks at a time.
Sensors that measure, for example, temperature and air quality are placed on ceilings, and others are placed in faucets to collect data on when they’re turned on. Sensors measure how often major appliances are used. There are passive infrared motion sensors and depth sensors to determine where people are in a room, too. Collectively, they provide a host of data, including the times of day residents cook and how often they make meals, Craddock notes.
Cameras are installed in the house’s main living areas, specifically focused on the couch in front of the television and the staircase to monitor sedentary behavior and physical activity. The system mines the video camera feeds, extracting information about, say, posture and whether the person is walking, exercising, or sitting, and for how long. To protect residents’ privacy, the video footage itself is not saved and cannot be seen by the researchers. The cameras and sensors are off-the-shelf, but the algorithms to capture data from them were developed by the SPHERE video experts, led by IEEE Senior Member Majid Mirmehdi.
Each person in the house wears a wristband with a number of motion sensors so the system can monitor who is engaging in which activities. The wearables, developed by Rob Piechcocki and IEEE Senior Member William Harwin’s research group, allow SPHERE to identify the participants. The wearables’ ultralow-power design means they can collect and stream data over a long period of time without being recharged. IEEE Member Bernard Stark is working on a way to wirelessly power the devices with energy harvested from, say, walking.
The data from all the sensors, cameras, and wearables are merged together to help answer questions about changes over time. Craddock’s researchers are now focusing on how to display the data so the occupants can learn about their health and behavior and then present the information to their doctors.
Data from the past may hold clues to a person’s present health: Is she sitting more than usual, drinking less water, or cooking or eating less? Is her mental state being affected by her sleep pattern or staying up too late at night? Seeing such information might prompt the person to alter a behavior in hopes of feeling better or to at least discuss it with a doctor.
Although the hardware will be the same in every home, the software can be tweaked to focus on a particular condition and the data that may affect it. “However, people and their families have evolving health conditions. They are not static,” Craddock says. “We are creating a platform that is flexible.”
The data can help the medical community better understand how certain conditions progress over time. But privacy concerns remain paramount, and occupants can adjust the system when they do not wish to have some, or any, data collected about them.
At this time, the system is not set up to advise users about what steps to take to improve their health based on the data. Nor is it designed to alert emergency workers if someone falls or needs help. However, such features are likely to be standard in smart homes in the future, so as to make the elderly and disabled safer in their homes.
“SPHERE’s role is to do adventurous research that companies and health care providers can’t do on their own,” Craddock says, given the staff, funding, and time required. “Once we can prove what the benefits of these technologies can be, others can follow in our footsteps.”
This article originally appeared in print as “A Home That Watches Over You.”
This article is part of our December 2015 special report on smart homes.