Eric Dishman is making a cup of tea-and his kitchen knows it. At Intel’s Proactive Health Research lab in Hillsboro, OR, tiny sensors monitor the researcher’s every move. Radio frequency identification tags and magnetic sensors discreetly affixed to mugs, a tea jar, and a kettle, plus switches that tell when cabinet doors are open or closed, track each tea-making step. A nearby computer makes sense of these signals; if Dishman pauses for too long, video clips on a television prompt him with what to do next.
It’s all part of a growing effort at Intel and other labs around the country to develop ways to help the elderly, and others who need assistance with everyday activities. Similar systems are in the works to monitor eating, sleeping, and medication habits in order to allow older people to live independently for longer. Researchers are even working on systems that analyze changes in behavioral patterns over time to provide early warning of aging diseases such as Alzheimer’s.
High-tech systems to monitor and assist the elderly are now becoming practical, thanks to the falling prices of sensors and processors, increasingly sophisticated software, and the wide availability of high-speed Internet access. They are also becoming increasingly attractive as a business; in the United States alone, the number of people over age 65 is expected to hit 70 million by 2030, doubling from 35 million in 2000, and similar increases are expected worldwide. “You’ve got many large technology companies like Intel suddenly noticing the aging demographics and asking, How will our future products fit into this space?’” says Dishman, who heads an Intel-led research consortium formed last year to develop monitoring technologies.
One of the simplest, nearer-term systems is under development by Honeywell Laboratories in Minneapolis, MN. The company is testing a home monitor for the elderly in seven assisted-living facilities in Minnesota and four homes in Florida. The Honeywell system starts with cheap, unobtrusive sensors set up around the home. Four to six motion detectors on the walls, plus a switch that detects when a pillbox is opened, are wired to a communications box in a closet which sends sensor information over the Internet to a processing station. There, software being developed by Honeywell compares what’s going on in the home-when a person gets out of bed, goes to the bathroom, and so forth-to patterns recorded during a calibration period.
The goal of the software is to glean a picture of the person’s daily activities. Motion in the bathroom and the opening of a pillbox, for example, would tell the computer that the person is taking medication. Activity in the kitchen would indicate the person is eating or drinking. Lack of these signals at certain times, or decreased activity overall, would suggest something is wrong; the computer would then make a telephone call with a simple reminder such as “take your pills.” The system could also alert caregivers, via either a call or an e-mail. Honeywell expects to sell the system in three to five years; while the prototype costs $5,000, the commercial version should cost less than $500, says Tom Plocher, the project’s leader.
The Intel consortium is developing even more sensitive ways to follow the activities of elderly people. Its research goes beyond motion detectors and pillbox sensors to include things like pressure sensors on an Alzheimer’s patient’s favorite chair, networks of cameras, and tiny radio tags embedded in household items and clothing that communicate with tag readers in floor mats, shelves, and walls. From the pattern of these signals, a computer can deduce what a person is doing and intervene-giving instructions over a networked television or bedside radio, or wirelessly alerting a caregiver. Dishman says Intel will install the first trial systems in the homes of two dozen Alzheimer’s patients by early next year.
Crucial to the most advanced systems is software. It’s one thing to get raw sensor information, but quite another to figure out what the person in the home is actually doing, says Misha Pavel, a biomedical engineer at the Oregon Health and Science University in Portland, OR. Working with Intel, Pavel’s team is developing artificial-intelligence algorithms that deduce a person’s intent by building a statistical hierarchy of possibilities-say, making tea, cooking, or doing dishes-that is based on past experience.
Longer term, software could even help detect disease. At the University of Rochester’s Center for Future Health, researchers are using networks of video cameras and powerful computers to detect changes in behavior and coordination signaling early-stage neurological disorders. In theory, a home system might detect the onset of Alzheimer’s or Parkinson’s disease before a patient deteriorates enough to seek a doctor’s help, says Philippe Fauchet, the center’s director.
One possible sign of early-stage Alzheimer’s that a monitoring system could detect: a person standing in the kitchen for a few minutes without doing anything. And to spot early Parkinson’s symptoms, the Rochester researchers are developing machine vision algorithms to extract the movements of a person’s arms, legs, and torso from video shot from multiple cameras in a room. This is the first step toward a software product that can detect very early Parkinson’s symptoms like decreased stride length and asymmetries in arm swinging. But turning these algorithms into practical systems will take time; Fauchet predicts commercialization will take a decade.
Health-care experts foresee no shortage of customers. Larry Minnix, president of the Washington, DC-based American Association of Homes and Services for the Aging, which represents 5,600 nursing homes and elder-care facilities, says consumers will pay handsomely for technologies that keep them or their aging parents independent, alleviate caregiver burnout, and improve nursing-home care. “Good care is expensive, but inadequate care is a lot more expensive,” he says. Big technology companies are betting he’s right, as they bankroll these systems’ transition from lab curiosities to demonstration models. “Two years from now, you will see many more trials of holistic home monitoring systems than exist today,” says Dishman. After all, these technologies are about improving the lives of the elderly-and developing new markets. Neither idea is likely to get old.
Sampling of Companies in Elderly Home-Monitoring COMPANY TECHNOLOGY GE Industrial Systems
(Plainville, CT) Low-cost wireless sensing system that caregivers can access Honeywell Laboratories
(Minneapolis, MN) Motion sensors and software that learns daily patterns of behavior in homes Intel Research
(Hillsboro, OR) Radio chips that track activity and software that detects cognitive decline Matsushita Electric Works
(Osaka, Japan) Interactive robot pets and advanced sensors to assist elderly nursing-home patients Motorola’s iDEN Subscriber Group (Plantation, FL) Smart cell phones that give reminders or directions and relay vital signs to caregivers
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.