Sunday, May 03, 2009, New Haven, Conn. — “Help, I’ve fallen and I can’t get up!” — so goes the catch line of a 1989 infomercial for a product designed as a simple alert system to aid people living alone.
The ad, though amateurish, reminds the viewer of what it could be like to be old, infirm and living alone. When an elderly person suddenly is incapacitated at home, they may be unable to get help, perhaps for hours or even days.
Two faculty members in Yale’s School of Engineering & Applied Science are pioneering the use of “smart cameras” to address these and related issues.
The average life expectancy in the United States is approximately 80 years and some predictions suggest that the number of people over the age of 65 will triple in the next 50 years. Increasingly, seniors are choosing to live independently — be it a matter of pride or cost. As a result, there is an increasing need for elder care services, and technology can help to fill gaps left by a limited care-giving workforce and dispersed families, say the Yale researchers.
Several products to address this need are now commercially available — camera systems and wearable devices. Eugenio Culurciello and Andreas Savvides are developing “smart cameras,” which are designed to enhance the prospects for independent living.
Culurciello, an assistant professor of electrical engineering, tackled the technology of fall detection. “Approximately one-third of individuals who are 65 and older fall each year,” he says. “While many falls do not result in injury, nearly 50% of non-injured fallers cannot get up without assistance, and the period of time they spend immobile often affects their health outcome.”
He has developed a surveillance system that definitively recognizes falls and automatically calls for help. “It’s simple, it’s inexpensive and it preserves the privacy and independence of the person being monitored,” says Culurciello.
The device requires only a “smart camera,” a high-speed camera with a microprocessor that analyzes rough outline images and distinguishes between patterns of motion. It is programmed to “know” the difference between someone who is sitting, bending, kneeling, walking — or falling. According to Culurciello, videos made from the fall detector information “easily distinguish a person who is falling versus a box falling off a counter or a cat jumping off a table.”
It is not just about the images themselves, he says — time is the key. Patterns that indicate a fall develop as a series of images over a particular time span, he explains. Once the processor registers the downward motion of a fall, it waits to see what happens during the next 30 seconds. If there is no motion upward, the alert goes out.
The device will call whatever support team the user chooses, Culurciello points out. The alert might go to family members, or an on-call home healthcare professional or a medical response team. “We see it as a practical and personalized way to coordinate families and teams for elder care,” he says.
Falls, however, are not the only concern. Dementia and depression are other issues that aging individuals often face and both may call for aspects of assisted living. Savvides, an associate professor of electrical engineering and computer science, applies smart-camera technology to evaluate patterns in the way people move around their living areas with an eye toward noticing telltale changes in behavior.