By Kanda Yaemboonruang
The elderly of our society are the age group with the highest risk of serious injuries from a fall, and this risk increases with age, the World Health Organization (WHO) stresses. Making matters worse, many seniors suffer such injuries alone without the knowledge of their families. As many countries around the world grapple with full-fledged aging populations, the issue will intensify.
According to a report by the National Economic and Social Development Board (NESDB), Thailand can expect to enter such a society in 2021 when the number of older people will rise to 13.1 million, accounting for 20 percent of the total population. This is a challenge for families, communities and the nation as more day-to-day care and assistance for the aging population will be required.
One hope to address the issue lies in technological advancement, specifically Artificial Intelligence (AI), which involves the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. Scientists are creatively applying the technology to many different sectors, such as healthcare and older persons’ well-being.
The Artificial Intelligence Center (AI Center) at the Asian Institute of Technology is working on ‘Elder Care’, an AI project in collaboration with the Faculty of Physical Therapy at Thailand’s Mahidol University. The main aim of Elder Care is to apply AI to monitor seniors’ activities, identify movement irregularities, estimate the possibility of falling, and alert caregivers before and during the fall. This can potentially reduce the impact of the fall and subsequent injuries as family members or caregivers can tend to the seniors in a timely manner and limit their medical and social costs.
Mr. Jednipat Moonrinta, a research associate of the Elder Care Project at the AI Center, explained that data collection and machine training is being conducted as we speak. The AI is fast learning about human movements and irregularities, he says. Normal human motions used for machine learning include things such as checking the time on one’s watch, clapping hands, and bending to pick something up while standing and sitting. Examples of human movement irregularities used for machine learning are collapsing from a heart attack, slipping, coming into contact with an object, and dizziness, to name some.
“We have invited a lot of volunteers to imitate those activities and simulate the various scenarios. We are collecting data from both normal and abnormal movements because we need the AI to learn what is normal and what is not, and to separate the two efficiently. This will allow the AI to recognize falls, collapses, and unstable movement, and alert caretakers or family members,” Mr. Jednipat explained.
In addition, AIT researchers have been collecting data on the movement and activities of a group of elderly participants by installing video cameras with an AI processing unit at sample houses in Bangkok and Chai-Nat provinces.
Supported by a research grant from Thailand’s National Broadcasting and Telecommunications Commission (NBTC), the project is expected to be completed in three months. The Elder Care Project intends to introduce the technology to families with their elderly members and later to provide the technology to nursing homes and senior living communities.