University of Waterloo too helps fix ‘emotional disconnect’ and support caregivers
University of Waterloo computer scientists have developed an emotionally responsive ‘virtual assistant’ to help individuals living with Alzheimer’s disease.
ACT@Home is an attempt to develop a home-based technology combining artificial intelligence with social psychological models, helping those with the disease to complete day-to-day tasks.
‘It was because he was living in the past and he thought he was in a war camp… The voice distressed him because he thought it was a call to arms.’-Jesse Hoey, Associate Professor at the University of Waterloo
Jesse Hoey, an Associate Professor at the University of Waterloo, has been working on assistive technologies for people with Alzheimer’s disease, and created this prototype to resolve the ’emotional disconnect’ he found between technology and the people who need it.
“We had a gentleman who was a war veteran and he was using on our systems,” said Hoey. In using speakers to give instructions for washing his hands, “he would get very upset.”
“We found that it was because he was living in the past and he thought he was in a war camp, and it was a male voice. The voice distressed him because he thought it was a call to arms.”
Emotional mapping
Alzheimer’s disease disrupts the memory of those afflicted, forgetting things ranging from when they should take medication, to memories and experiences.
As it progresses, the disease leads to problems with reasoning and behaviour, and impairs a person’s ability to live independently.
To develop an emotionally aware prototype, Hoey interviewed individuals with Alzheimer’s disease and their caregivers, to better understand the interactions and communication that goes on in that relationship.
“We’re trying to build a map of the person’s identity, so who are the different people that they have been in the past and who they are right now,” he said.
The technology will use cues like tone of voice and facial expression, and use biographical information, producing an appropriate reaction for the individual’s needs.
“Maybe a job they had in the past, and current situational identities like a patient,” he said. “If we build a computational map of those things, and then use that to make our best guess as to what interaction will work best.
Helping at home
Hoey hopes that the technology will allow those with Alzheimer’s to live longer and better at home, and help ease the stress of an informal care partner like a spouse, child or friend.
“We’re trying to provide tools for these people to use that will help them to help this person, and take away this burden from them so they can do a little less,” he said.
Home care can be a challenge for the families of those with Alzheimer’s, especially in areas like hygiene “where independence is highly valued.”
While the emotional intelligence of the technology will take a couple of years, Hoey is hopeful that in the future the assistant could be applied beyond Alzheimer’s patients.
“Our goal is to build technology that would help people live at home for longer,” he said, and that could potentially expand to those with Down Syndrome, types of dementia , or traumatic brain injury.
Clare Bonnyman 