Advertisement

UBC student builds AI voice-controller for brother's wheelchair

An engineering physics student at the University of British Columbia has invented a voice-operated wheelchair for his brother, who suffers from Duchenne muscular dystrophy.

Michael Ko doesn't have a background in programming or electrical engineering. Instead, Ko says he relied mainly on YouTube videos to build the chair's controller, which responds to his brother's voice.

"He's my brother. I think that's the main driving force behind it," Ko said.

"I always see him as a guy with perseverance, never giving up. And that's the person I want to be as well."

Michael's brother Daniel, 28, has struggled with the muscle-wasting disease for years. He began using a wheelchair when he was eight.

CBC
CBC

Seeing his brother struggle pushed Michael to do more to help.

Daniel Ko used to use a voice-activated program, Google Assistant, to help get around, but in 2017 surgery damaged his vocal chords and the software was no longer able to recognize his commands.

Most voice-recognition software can no longer interpret Daniel's speech, which is often interrupted mid-word by having to take a breath.

"That's when I decided I really wanted to help him," said Michael Ko.

To do so, Michael decided to build his own voice recognition software, which could then help operate Daniel's wheelchair.

CBC
CBC

It has taken Michael a lot of trial and error — and a few exploding electrical components — to create the right device. But now Daniel can operate his wheelchair with a few simple commands that are easier to pronounce.

"At first I thought it was impossible that an engineer would be able to do this but now were practicing and the voice recognition is getting better and better," Daniel Ko said.

Michael called his device Ava, named after one of the characters in the two brothers' favourite movie, WALL-E.

"I wanted to base it off of that because the robots in WALL-E, they help humanity," Michael said. "I wanted to transfer that same message that these AI [artificial intelligence] devices ... can really help us in the future as well."

As well as learning about software and electrical circuits, Michael says the project has taught him other valuable lessons as well.

"What it's really taught me is if you have a passion for something, if you have a drive for something then nothing is really impossible," he said.