NewsLocal News

Actions

Morgan State students test autonomous wheelchair that uses AI

BWI Autonomous Wheelchair 2.png
Posted
and last updated

BALTIMORE — According to the state, BWI Thurgood Marshall Airport saw over 2.6 million passengers pass through their gates in July. Among them, at least 63,000 passengers needed assistance.

To help them better, researchers at Morgan State are developing an autonomous wheelchair. It will help passengers with disabilities independently move through the airport.​

"I'm legally blind," said Julie Terrill.

Terrill arrived Tuesday morning to visit family for the first time in Baltimore. She's pretty independent, but unfamiliar airport, unfamiliar obstacles.

"If it's a larger airport I have a wheelchair assist," Terrill explained.

She tells me its easier than stopping constantly to ask people for directions. But the service isn't always efficient.

"I have to wait for a long time for someone to be available," said Terrill.

Now, transportation and electrical engineering students from Morgan State are testing their autonomous wheelchair to deliver passengers from parking lot to gate at BWI. They've been testing here for over a year. Julie likes the idea.

BWI Autonomous Wheelchair.png

"I think that's fantastic. I would try that. It's a little daunting because I feel someone should be able to see," Terrill said.

Enter the tech. The eyes take the form of QR coded tracks and LiDar to navigate the busy halls.

Dr. Kofi Nyarko is in charge of the engineering side of this project.He's the Director of The Center for Equitable Artificial Intelligence and Machine Leaning Systems. He explains how the LiDar works.

"It goes from one place to another based on an internal map, sensing obstacles along the way," said Dr. Nyarko.

Dr. Mansoureh Jeihaniis professor and directorof The National Transportation Centerat Morgan State.​

"So, it makes you independent and mobile," Dr. Jeihani said.

Dr. Jeihani directs the transportation engineering side of things. She says this is the third phase of the project. Tests in the campus lab were successful, then came the real world.

"When we came here, there were so many different details, different lighting, shadowing, different texture and color of the floor and reflection that comes to the camera and all that," she explained.

Again, technology came to the rescue. Dr. Nyarko explains further.

"We do use machine learning, which is a form of artificial intelligence. And we have used the system's ability to collect lots of examples of what the guides look like under varying conditions. And trained itself to distinguish what is the guideline and not the guideline," he said.

From here the sky is the limit. And Dr. Jeihani hopes to see others use this device elsewhere.

"We started from the airport but it could be used in hospitals, in museums, any large buildings that is confusing where to go," said Dr. Jeihani.

And as for Julie, she's says lets go for it.

"Any help I can get, I appreciate. I can't imagine having to have done this fifteen years ago," Terrill said.

The next phase of the project is to practice full autonomous navigation in the airport's busiest locations. So, next time you're flying through, watch out, an AI wheeler may pass you by.