Smart monitor for elderly

Sensors and computers are coming close to doing what good neighbours used to do - checking that elderly people living alone have turned the TV off at night and cleared the letterbox daily, and calling for help if they notice a problem.

A team at Massey University's school of engineering and advanced technology is completing refinements to its SAM (selective activity monitoring system) that will send a text message alarm if a home's occupant doesn't do what they normally do.

The prototype has already won an award, the Digital Signal Processing creative design contest, judged in Taiwan in November last year.

The challenge now, says associate professor Subhas Mukhopadhyay, is to find a volunteer to test the smart digital home system in a real situation, and an aged care housing or home support group to get involved.

"Quite often we are appalled by news headlines such as 'Elderly man lay dead for days in his home' or 'Woman found starved in flat'."

But the reality is that more and more elderly people are choosing to live privately and independently even though they and their families know there are risks, he says.

SAM aims to use microtechnology to manage the risks non-invasively.

The SAM technology has been developed to monitor use of household appliances and call for help if use patterns change. It's launched graduate student Michael Sutherland into a career with Fisher and Paykell's medical division, and now provides doctoral student Anuroop Gaddam with a specialist topic.

SAM sensors can be fitted inside wall sockets where the toaster or electric kettle is plugged in, or to the television, and subject to further developments, to the bed, the letterbox, or the hot water cylinder.

The supporting software picks up changes outside what's normal for the occupant. It can be programmed to raise the alarm if the kettle hasn't been turned on to boil water for a hot drink by a certain time in the morning, if the television isn't turned on and off again in the evening, or the person doesn't get out of bed in the morning.

Dr Mukhopadhyay says the system will be more acceptable to elderly people who value their privacy than movement sensors or blatantly intrusive cameras.

There is no one person constantly monitoring the occupant's movements, just the software, that will only text for help if patterns change and the audible alarm isn't over- ridden.

Another advantage is it doesn't depend on a person being conscious, recognising a problem, and being physically able to push a button for help. The text will be launched even if the person has fallen and is unconscious.

It's then up to the caregiver or family member on the receiving end of the text to respond.

The sensor units look likely to cost $40 to $50 at the moment, but that price would come down with mass production.

Computer science professor Hans Guesgen is already working on ways to further refine the system to make sense of complex behaviours that are demonstrated by elderly people experiencing some degree of cognitive impairment.

The "ambient intelligence" would be able to alert caregivers to potentially dangerous situations - like forgetting to turn electrical appliances off. He is also working on ways to help such people interact with computerised systems, perhaps through gestures.