Dr. Goldie Nejat at the University of Toronto sees a day when socially assistive robots will help those living with cognitive impairments to complete everyday tasks.

“The idea is to promote engagement in the task and engagement with the robot as well,” says Nejat, director of U of T’s Autonomous Systems and Biomechatronics Laboratory. 

Nejat and her team have conducted several short-term studies using social robots in long-term care homes and seniors’ residences in Ontario. Now, the first long-term user study — done in partnership with AGE-WELL and Dr. François Michaud, founding director of the Interdisciplinary Institute for Technological Innovation (3IT) at Université de Sherbrooke — is currently underway with seniors in a variety of settings in Quebec and Ontario. 

Dr. Goldie Nejat

According to Nejat, the project is the first of its kind in Canada. She hopes to expand her research across the country to see if there are any differences in effectiveness as her team adapts the robots to different users in different settings — be it a long-term care home, retirement home or private residence.

The team is continually dialing in the robots’ ability to read social intonation and prompt residents to interact as the studies return more data, she adds.

“We’re looking at how the robots can help older adults to complete essential and important activities of daily living,” Nejat says. 

Nejat and her team at U of T’s Autonomous Systems and Biomechatronics Laboratory have built and experimented with several unique robots.

The robots, such as Casper the Friendly Robot (who was used in a home study for assisted living), or Tangy, the socially assistive robot who successfully facilitated a bingo game in a lab experiment, share many common traits when it comes to appearance.

Photo Courtesy of Dr. Goldie Nejat.

They are around four feet tall, on wheels and have screens in their chests. Some of the units, like Casper, have human-like faces, but they are a little more cartoonish and their mouths and eyebrows are lights that pulsate rapidly to express emotion.

Nejat says the robots learn to recognize user affect and respond accordingly with their own emotions to promote engagement.

They can remember individual users for personalized interaction and provide caregivers with alerts if there is a need for human intervention.

They can also recognize if a person is transitioning from one phase of dementia to another. 

“They may start to have physical difficulties like manipulating a fork, for example, and the robot can recognize this over time and alert caregivers to this new difficulty, bringing it to their attention,” Nejat says.

“So, we are really looking to personalize the experience and give users that person-centred care.”