WRITTEN BY EMILY BAER
As auto manufacturers continue to develop and refine autonomous vehicle (AV) technologies, it often feels like the future of driving is out of our hands. Increasingly, driverless cars are being tested in urban centers and tech hubs; these practical experiments are teaching us, in real time, what these innovations can offer communities in terms of safety improvements, efficiencies, and opportunities for accessibility across diverse populations. But too often there’s a key element missing: humans. For researchers at the University of Maine’s VEMI Lab, this presents an exciting opportunity for creative thinking and adaptive
problem solving.
“A lot of our research is based on the idea that yes, this technology is great and the safety improvements are fantastic,” explains VEMI director Dr. Richard Corey.
“But a lot of times what is being developed is done without a sense of what a real human may or may not need within it.”
VEMI’s focus on ‘human-vehicle collaboration’ addresses those gaps and their work is making a tangible impact on how AI technologies are developed. Increasingly, their research has caught the attention of industry insiders, policy experts, and the federal government.
Founded in 2008, VEMI has long been devoted to exploring the intersection of human behavior and technology to solve unmet challenges. By leveraging diverse, interdisciplinary perspectives they help develop solutions to technologies that prioritize accessibility for all users.
In 2022, VEMI’s Autonomous Vehicle Research (AVR) Group participated in the U.S. Department of Transportation’s Inclusive Design Challenge (IDC). UMaine researchers developed “Ava,” an Autonomous Vehicle Assistant, and ultimately claimed third place in the competition, receiving national recognition and a $300,000 award.
In the years since, the AVR group has focused on understanding how autonomous vehicles need to accommodate and interact with their human users, building on relationships that were catalyzed through the competition. Corey and his team see a unique opportunity for Maine to lead research in this field.
“When we started this work, the vast majority of advancements in the field were based on engineering developments,” says Postdoctoral Research Associate Paul Fink. “We are interested in how the technology is being designed to help those who are poised to benefit from it the most. We are focused on the human side of things.”
The AVR group is particularly interested in exploring autonomous vehicle applications in rural communities, for low or no vision users, and for those with limited mobility. “We want to explore how we make these systems work well for all people, but especially those who are most excited to have opportunities, autonomy, and mobility that they otherwise don’t have.”
As a rural state with an aging population, Maine is ideally suited for autonomous vehicle testing. “AV has the potential to provide independence for people, particularly older folks, who want to live on their own but don’t want to live in urban centers. Unfortunately a lot of the testing right now is happening in cities,” says Fink.
At the same time, Maine weather presents ample opportunities for refinement. Many of the cities where testing occurs today are in warm, arid climes but manufacturers will need to contend with diverse weather patterns and events. In that sense, Maine’s four seasons are an incredible asset. “You might not think that our weather would have much of an impact, but then suddenly you realize that computer vision gets a little squirrely when you have a bunch of yellow leaves and dotted yellow road lines,” explains Corey. “Not to mention snow. Snow can really throw a computer off.”
Few other states are invested in exploring these practical challenges, but there is increasing urgency to do so. The AVR group at UMaine is at the forefront of this research, with growing collaborations with industry innovators and increased visibility in the field. “This is the place to do this work,” says Corey. “This is where
this research needs to happen.”
