AI

John Robinson was born without the extensions of his arms or legs. As a child, and an adult, he rejected wearing prostheses. They worked, but they were uncomfortable, and he never felt like himself with them on. And that’s all Robinson really wanted to be, just himself.

It’s why he understands the frustrations many other people with disabilities face in trying to get people to see who they really are – especially when it comes to looking for employment. And why his organization, Our Ability, is among seven new recipients of Microsoft’s AI for Accessibility grants to people using AI-powered technology to make the world a more inclusive place.

In 2018, when the $25 million program was announced, nine organizations were given grants to work on a variety of projects, some from scratch, some already underway.

The new 2019 grantees are: University of California, Berkeley; Massachusetts Eye and Ear, a teaching hospital of Harvard Medical School; Voiceitt in Israel; Birmingham City University in the United Kingdom; University of Sydney in Australia; Pison Technology of Boston; and Our Ability, of Glenmont, New York.

Their projects may differ, but the people behind them share a passion for how technology can improve the lives of their fellow human beings.

“What stands out the most about this round of grantees is how so many of them are taking standard AI capabilities, like a chatbot or data collection, and truly revolutionizing the value of technology in typical scenarios for a person with a disability like finding a job, being able to use a computer mouse or anticipating a seizure,” says Mary Bellard, Microsoft senior accessibility architect.

The one-year grants provide use of the Azure AI platform through Azure compute credits and can also include Azure compute credits plus engineering-related costs. AI for Accessibility has three focus areas: communication and connection; employment; and daily life.

Robinson of Our Ability knows about all of those. But it was his employment journey after college that always remained with him. He went through a discouraging, 4-1/2-year search to find a job – the right job.

He vowed, someday, to find a way to improve the process for others who have disabilities. “You want to be seen as the person you are, in total, not just the shell that you are on the outside,” he says.

Robinson founded Our Ability in 2011 to bring businesses with employment opportunities together with people with disabilities looking for jobs.

Now, with the AI for Accessibility grant, and working with students from Syracuse University, Our Ability wants to create an accessible and intuitive AI-powered chatbot to help businesses find workers, and to help people with disabilities find employment that’s meaningful to them.AI

Every day, Dexter Ang was growing more frustrated as he watched his mother deal with the indignities of ALS, amyotrophic lateral sclerosis. She was diagnosed with it in 2014.

Ang spent months meeting with experts and poring over information about existing technologies to see if there were any that could help his mother.

He wanted to learn more, returned to MIT as a graduate student and co-founded Pison Technology. The company, one of the AI for Accessibility grantees, is developing a nerve-sensing wearable, similar in appearance to a watch, to control digital devices using small, micro-movements of the hands and arms.

The device, which is now undergoing testing, can help improve communication for people with other neuromuscular disabilities, including multiple sclerosis, Ang says. He wants it to be made widely available, around the world, at a low cost, and easy to purchase online. He believes his mother, who died in 2015, would be proud of the work he and his team have done.AI

For people with epilepsy, one of the biggest dangers is having a seizure while driving. In some places, patients need to prove they’ve been seizure-free for a year in order to be allowed behind the wheel, which can create psychological and economic stress of its own.

It’s a problem that got Omid Kavehei thinking: What if there was a way to warn drivers who have epilepsy that a seizure could be coming, so that they have time to safely pull off the road?

Kavehei, a senior lecturer at the Faculty of Engineering and Information Technologies, and his colleagues at the University of Sydney – also a new AI for Accessibility grantee – are working on a potential way to help. They have been using deep learning to develop an analytical tool that can read a person’s electroencephalogram (EEG) data via a wearable cap, then communicate that data back and forth to the cloud to provide seizure monitoring and alerts.

Kavehei and his colleagues want to first test a wearable cap on epilepsy patients using driving simulations. They will leverage Azure Machine Learning to attempt to predict seizures from human signals.

The research being done by all of the AI for Accessibility grantees “is an important step in scaling accessible technology across the globe,” says Bellard of Microsoft. “People are looking for products or services to make things easier and AI might be able to help.”