AI robots deployed to fill void in senior care. But can they?

The COVID-19 pandemic has been particularly cruel to older people. Digitally estranged, those living alone were virtually cut off from the world as social distancing disrupted operations of senior centers and other community services.

To fill this void in senior care, authorities and institutions have deployed companion robots.

These care machines with artificial intelligence have been billed as an example of the pandemic accelerating the deployment of cutting-edge technology. They are intended to provide emergency monitoring for older adults living alone, and help ease their loneliness, developers say.

Meet Hyodol and others

“Hyodol” is one of the most well-known AI companion robots for seniors here, developed by a startup of the same name.

Since its launch in 2018, more than 5,400 units of the AI-powered stuffed toys have been supplied to 114 local governments including Jongno-gu in Seoul and 250 institutions nationwide, according to the Seoul-based company.

With the convergence of sensors, Internet of Things and machine learning technologies, the service robot notifies the guardian of a user when no movement is detected for a certain period of time. It sends emergency text messages and phone calls when a user presses and holds its hand for more than three seconds, a gesture that can be custom set for such situations.

Standing 35 centimeters tall, the doll is also loaded with 9 types of entertainment programs, including sets of quizzes and stories. It can play more than 3,000 songs, with genres ranging from classical to trot -- an old Korean pop genre loved by seniors.

Among Hyodol’s features, senior users like the talking alarm function the most, as it reminds them to take medicine, said Lee Yae-seul, a Hyodol official.

The doll is designed to talk when touched by the user and in accordance with customizable time settings such as meal times.

“We want an emotional exchange (between users and Hyodol) so that users feel like they are being cared for and that there is someone by their side,” the Hyodol official said.

Another caregiving companion robot adopted by local governments including Haenam County Office in South Jeolla Province is Dasomi, developed by Wonderful Platform.

Dasomi looks like a barcode reader with a screen, and has functions similar to Hyodol. The screen, which can play dementia prevention gymnastics videos on demand, works as a video conferencing device that connects its users with family members or government officials, allowing them to remotely check on the user.

Seoul-based Mr. Mind also launched its version of fluffy AI companion robots, which the firm says monitors and tracks mental and physical health by storing and analyzing what users say.

How far have the robots come?

A study published in the Journal of the Korean Gerontological Society last year found that users were “moved and strengthened” by the words of encouragement from Hyodol.

The authors, including Lee Eun-kyoung, a professor at the University of North Carolina School of Social Work, concluded that the robot helped research participants -- 12 people whose average age is 82 and have used the robot for 18 months on average – improve their daily activity and depression levels.

It also found that the users felt discomfort when Hyodol failed to understand their words, and when their hopes for a two-way communication and building bonds were not met.

Another study conducted by Song Moon-sun, a doctoral student at Ewha Womans University, found that users started to experience a personification process of Hyodol around three months of use.

“In the first encounter, female participants were more likely to perceive Hyodol as a ‘doll,’ while men experienced anthropomorphic disorders as they recognized it as a ‘machine,’ showing dissatisfaction with the low intellectual level of the robot and simplicity of their conversations with it,” Song said.

In another study published by the Society of Digital Policy & Management in February, Suwon Women’s University’s Nursing Professor Kim Yeun-mi found a similar effectiveness in the use of robots for the elderly in one-person households.

“If such non-face-to-face care technology is introduced to the elderly care field in line with ‘living with COVID-19,’ it can be expected to contribute to cognitive function training and reduction of depression among the elderly,” Kim noted in the report. The research was conducted on 38 people, and included an experimental and control group.

Lee Kyoung-jun, a business management professor at Kyung Hee University, was skeptical of an AI companion robot like Hyodol satisfying seniors’ needs.

He said even tech behemoths’ virtual assistant programs, including Apple’s Siri, Google’s Assistant and Samsung’s Bixby, have not reached the point where they can communicate with users like human beings.

“Giving physical care (through a robot) may be easy, but emotional care is much more difficult,” he said.

For similar reasons, Oh Yoon-seon, an assistant professor at the Department of Electronic Engineering at Hanyang University in Seoul, said it is better for senior care technology to focus on medical monitoring, rather than the interaction capabilities.

“Especially for elderly users with mild dementia, it is important that the robot is capable of detecting their activities in the event of a collapse or the user leaving home without a companion,” she said.

Deep learning technology may also be needed to facilitate human-computer communication, as some seniors are unfamiliar with using digital gadgets with complex interfaces, she added.

By Park Han-na

Asia News Network: The Nation (Thailand), The Korea Herald, The Straits Times (Singapore), China Daily,  Jakarta Post, The Star and Sin Chew Daily (Malaysia), The Statesman (India), Philippine Daily Inquirer, Yomiuri Shimbun and The Japan News, Gogo Mongolia,  Dawn (Pakistan),  The Island (Sri Lanka), Kuensel (Bhutan), Kathmandu Post (Nepal), Daily Star (Bangladesh), Eleven Media (Myanmar), the Phnom Penh Post and Rasmei Kampuchea (Cambodia), The Borneo Bulletin (Brunei), Vietnam News, and Vientiane Times (Laos).