UCSD Jacobs School of Engineering

Students recognized for smart wheelchair research at CHI

Goto Flickr
UC San Diego students at CHI in Scotland. 

San Diego, Calif., June 27, 2019 -- They interviewed 70 people from seven countries by email, phone and Skype. They researched medical conditions. They investigated Lidar, camera sensors and other off-the-shelf technology.

And at the Conference on Human Factors in Computing Systems (CHI) in Scotland, a team of undergraduates from UC San Diego won third place for this work in the student research competition. Their research aimed to understand the needs of wheelchair users, and develop an affordable, smart wheelchair kit with multiple levels of autonomy that would be useful to them. 

Based on these interviews—some of which were written, and several of which were in-depth video interviews over the course of several conversations—the students decided that a kit to turn a standard electric wheelchair into one with multiple levels of autonomy and voice control would most benefit people with Myalgic Encephalomyelitis (ME), more commonly known as chronic fatigue syndrome.

“One woman explained that on a good day she can do so much, but then on a bad day she really can’t do anything, and needs help with things like eating,” said Ryan Lin, a computer engineering student and member of the team. “I think that’s what really got us to try and address multiple levels of autonomy. We want to make her good days the best and most independent, and help her on her bad days.”

Their goal was to develop a kit that contained the off-the-shelf technologies needed to turn a standard wheelchair into one with two levels of autonomy, all at an affordable price and without needing complex engineering skills.

“Our goal is to make a kit that anyone could build with just basic skills-you don’t need to be an engineer or anything,” said Jesus Fausto, an electrical engineering student. Their goal is for the add-on kit to cost no more than $1,000.

The lower level autonomy function—level two autonomy-- could be used on a day the user is feeling strong. At this level of autonomy, the user is totally in control of the wheelchair, but the Lidar and camera sensors on the chair are still active and will sense and avoid objects in its path. Several people surveyed mentioned that their chairs had nicked door frames and damaged items in their home, so this level of autonomy is meant to avoid those accidents by braking or moving before hitting an object.

Goto Flickr
A prototype of a standard wheelchair equipped with the team's autonomy kit. 

On a bad day, users could take advantage of the higher level of autonomy-- level four autonomy-- whereby the user simply says where they want to be taken, and the chair autonomously gets them there.

“With the depth camera we create a 3D map of the house,” explained Fausto. “It’s automatically done—you hook up the chair, drive it around your home once, and it creates a map of the house.”

This functionality is in a prototype phase, but uses Amazon Alexa as the natural language component. It can currently only be used indoors, where the chair can build a map of the space. The students are still working on integrating the brain of this system (the natural language cues) in sync with the body (the movement of the wheelchair), though both components work separately.

Wheelchairs with similar capabilities exist. But the catch is they’re extremely expensive, often costing tens of thousands of dollars. The students’ kit uses off-the-shelf commercial products including Lidar sensors, infrared sensors, depth cameras and voice assistants like Amazon’s Alexa to turn a standard wheelchair into a smart one for a fraction of the cost.

Ultimately, the team envisions creating different versions of the kit based on what products are available in different parts of the world. The project will also be open source to enable maximum reach.

“That was my role on the paper, I interviewed a lot of people in Sri Lanka, India, China, Japan, and Australia and New Zealand to understand what economic lines are they comfortable with and what wheelchairs cost in each of these different countries, because there are varying prices,” said Samarth Aggarwal, a cognitive science and math double major. “One of the goals is also to find out how to manufacture this on every economic line in every country, because if you make it in the US, the economic cost is much higher and people in other countries can’t afford it. So we’re working to make sure all this tech is reproducible with locally sourced versions.”

The project was an interdisciplinary collaboration, with students and support from electrical and computer engineering, computer science, nanoengineering, cognitive science, and mechanical and aerospace engineering. It is led by Jack Silberman, a lecturer in the Jacobs School of Engineering and the commercial quality assurance manager for Canada and Latin America with Abbott Diabetes Care.

Roughly 20 students contributed to the project over the course of a year. Five students presented the research at CHI: Isabella Gomez Torres, Gaurav Parmar, Samarth Aggarwal, Nathaniel Mansur, and Alec Guthrie. More information here: https://smartwheelchair.eng.ucsd.edu/

 

Print News Release