Overcoming obstacles: AI-powered app flags mobility barriers for people with disabilities

Rehabilitation medicine, engineering researchers team up for pilot project to help navigate accessibility challenges on U of A campus.

Arne Andres has spent a lifetime confronting barriers to mobility.

Growing up in the Philippines, he was struck by a childhood bout of polio that left him in leg braces. Friends and family had to carry him in a little wooden box to get from the street to his classroom.

As a computer science instructor later in life, he struggled to climb the stairs in the building where he taught.

Accessibility was much better when he immigrated to Canada in 2001, but in a cruel stroke of bad luck, he was rear-ended in a car accident in 2015, leaving him severely disabled. He has been using a motorized wheelchair ever since.

One day while recovering in Edmonton’s Glenrose Rehabilitation Hospital, Andres saw a flyer advertising an open house at the University of Alberta’s Faculty of Rehabilitation Medicine. That’s where he met the Click&Push Accessibility research team that would change his life.

Click&Push was developing a free, community-sourced navigation app called The Atlas to help users access public spaces more easily, using a crowd-sourced map to show them in real time the barriers they should avoid — such as elevators out of order, construction sites or other obstacles.

Much like a traffic report, the app would allow people to avoid unnecessary frustration by planning their day before leaving the house.

Andres immediately signed up to help test the app and ended up becoming a director at the startup. That led him into a second master’s degree, this time a cross-disciplinary study in rehabilitation medicine and engineering that uses artificial intelligence — machine and deep learning — to identify and classify mobility barriers around the U of A’s North Campus.