lderly people who want to remain in their homes for as long as they can may one day get help from an unlikely source: the homes themselves.

Academic and corporate researchers are developing systems that could be installed in homes to monitor the occupants' health or provide unobtrusive memory aids. These "smart" homes would use the growing power of computer networks and sensors to help the elderly avoid or postpone institutional care.

Inexpensive sensors that detect sound or vibration (and send that information to a computer) could, for example, monitor footfalls to analyze the pace at which someone climbed the stairs or moved between rooms. If a pattern changed — perhaps because of a sudden weakness in gait — the system could alert friends or family living elsewhere.

Other systems could provide low-key help for lapses in memory by providing reminders for the names of people or objects or documenting tasks with a camera so people who were interrupted, say, when cooking could return to the stove, examine the images and take up where they left off. The technology could also remind people to take medications, eat or drink water.

With the market for such services expected to be huge, technology to make it possible is in the works at the Georgia Institute of Technology, the University of Rochester, the Massachusetts Institute of Technology and elsewhere.

"The elderly are going to be an enormous slice of the total population," said Dr. Jim Larson, a computer scientist at the Intel Architecture Labs in Hillsboro, Ore. Intel sponsors some smart-home research projects directed toward older people.

Dr. Larson said that research into how to build smart homes would result in affordable products within three to five years. "This technology will make a tremendous difference," he said, "if it means that people can stay in their homes even a few months longer, or 18 months or two years."

Like any kind of electronic monitoring, the use of sensors embedded in a home or worn by people — things like microphones and cameras — raises issues of privacy. But some experts say older people will be willing to trade some privacy for an increased sense of security. There are also concerns about how much such systems will cost and how easy they will be to use.

More than 34 million people in the United States are 65 or older, and technology is already being used to keep some of them safer. Some people wear emergency alarms on pendants, and some retirement homes use door sensors that issue a warning if a resident has not left an apartment for meals. And on the horizon is the baby boomer generation: 76 million people born between 1946 and 1964, the largest group ever in the United States to head toward retirement.

Taking care of aging baby boomers will be a strain on younger members of society, said Dr. Alex Pentland, a baby boomer himself and the academic head of M.I.T.'s Media Laboratory. "Folks like me will be old and nobody will take care of us," he said. "We're up against a demographic — we have to invent our way out of it."

Dr. Elizabeth Mynatt, a computer science professor at Georgia Tech, said it was important to take families' concerns into account when designing technology for the elderly.

"Parents are often forced to move out of their homes for emotional reasons rather than logical ones," Dr. Mynatt said. Even when parents are doing well living alone, she found, many children who are unable to check on them daily become so worried that this anxiety finally precipitates the parents' move to assisted-living communities or nursing homes.

Dr. Mynatt, who conducts much of her research in Atlanta at a two-story, 5,000- square-foot house that was built by the state as a showcase for intelligent technology, decided to address this problem with memory aids and a monitoring system. The memory aids could prompt someone to take medication or finish a task, and the monitoring system could provide a digital portrait of the occupant's level of activity to relatives, even if they were far away.

The team is working on displaying information on an Internet device that would look like a picture frame containing a photo of the elderly person; for now, the display is shown on a computer monitor. Around the photo are butterfly or flower icons that grow larger or smaller to indicate the person's activity level. Each elderly person would decide what type of information would be transmitted to children or other relatives or friends.

"I want an interface that will reassure adult children that their parents are moving around normally," Dr. Mynatt said, "but not in the form of graphs and charts. It has to be easy to look at and understand."

The first versions of Dr. Mynatt's display and tracking system, called a digital family portrait, were based on data obtained by interviewing the elderly person each day. Now Dr. Mynatt is working with a team of scientists who have mounted sensors throughout the home to create the occupant's digital portrait.

A live demonstration of part of the digital portrait system was shown at the Association for Computing Machinery's recent conference on technologies of the future, held last month in San Jose, Calif. Movement was tracked by small, inexpensive radio- frequency tags worn by people in the house; the signals were picked up by antennas hidden, for instance, behind the living room chair. "Then people can see on the computer monitor what it looks like when the inhabitants move around: is it normal? Or is something unusual happening?" Dr. Mynatt explained at the exhibition.

The cost for a radio-frequency system for one room is about $800 to $1,000, she estimated, with an additional cost of about $400 for the computer and the display.

At the University of Rochester, researchers are preparing a test living room, kitchen and bedroom where they will develop technology for the elderly. Among the projects is one to develop software that would use pressure sensors and cameras to detect changes in gait.

Dr. Alice Pentland, a professor and medical director of the Center for Future Health, a research laboratory at the University of Rochester's Strong Memorial Hospital, said: "Our notion is to provide tools that people don't have to have a doctor's permission to buy. We are all trying to avoid seeing the doctor." Dr. Alice Pentland and Dr. Alex Pentland are siblings, and they collaborate on some research projects.

To track occupants in smart homes, most of the current projects use small portable tags or badges that send and receive radio waves or infrared light. They are inexpensive alternatives to camera systems, which some say are not yet technically adequate to do the job. The tags can be attached to things like lapel pins, key chains and medicine bottles.

At the M.I.T. Media Lab, for example, Dr. Alex Pentland and his colleagues, including Steven Schwartz and Richard DeVaul, are developing wearable computers, contained in special lightweight vests, that would use an infrared system of tags and detectors to identify places, people and objects and to provide medication reminders. The information could be displayed on a tiny monitor that was part of a pair of eyeglasses.

Dr. Alex Pentland likes the idea of using cameras, he said, but "they are not ready for prime time."

Cameras are going to rule one day at the Georgia Tech house, though, staff members there say. Dr. Irfan A. Essa, a computer science professor at Georgia Tech, is one of the people building a tracking system, based on video cameras, that will one day replace radio frequency tags. "We can locate where the person is," Dr. Essa said, "and make a first-level guess at where this person is heading using the optical sensors."

People are scared by the word camera because it sounds as if Big Brother is watching, he said. "But our intention isn't to spy," he said. "We want to use the system for early warning."

At the M.I.T. architecture department, Dr. Stephen S. Intille, a research scientist, is also concerned about the use of cameras and privacy fears as he helps design the M.I.T. Home of the Future, which will probably use camera monitoring. "I think people will change their minds as they become aware of the benefits of aging in their homes," Dr. Intille said.

Dr. Aaron Bobick, a computer science professor at Georgia Tech, predicted that camera-equipped houses would be practical. "If it costs, say, $5,000 to fit up your house, you might pay that to stay there." And it is possible that such services may someday be offered for monthly fees.

Leon Harper, a specialist at the AARP on housing for the elderly, is not as concerned with privacy issues as he is with design issues. "The technology must be comfortable, convenient, flexible and easy to use," he said.

"Smart houses are a good idea," he added, "but only if the houses are designed to accommodate the way people actually live — and the more unobtrusively, the better."