��������

KALAMAZOO, Mich.—Autonomous technology is driving the future of transportation worldwide. But what good are self-driving cars if they can only operate in perfect weather? A team at �������� Michigan University is leading an effort to develop software that will keep systems running in all conditions.

Dr. Zach Asher, assistant professor of mechanical and aerospace engineering and director of ��������'s Energy Efficient and Autonomous Vehicles Lab, and doctoral student Nick Goberville were awarded a grant from the Michigan Translational Research and Commercialization—MTRAC—Innovation Hub for Advanced Transportation. The grant will help bolster research for a new business venture they've launched through �������� called .

"There's a lot to be done here. There are lots of different types of weather," says Asher. "You can have extremely dark conditions, which are challenging for driver's assistance and autonomous systems. You can have heavy snow cover, light snow or actual snowing conditions. You have rain, heavy rain, fog. Each of these may result in a little bit of nuance. On top of that, we have all these different types of vehicles and levels of autonomy, so we've really got our hands full here."

Asher and Goberville plan to use the MTRAC funding to develop basic software that informs technology on how to drive in bad weather. They’ll collect data from a vehicle outfitted with a camera and sensors.

"Once the video frames are received from the camera, we are basically applying various machine learning and computer vision techniques to process the image which provides a safe, drivable region for the vehicle to traverse," says Goberville. The team expects to have a viable product to debut in August 2021.

Their research is among a select group of projects identified by the organization that aim to address future or poorly met market needs, offering ways to increase the efficiency, safety and sustainability of moving people and goods. Through MTRAC, each team of researchers receives mentorship support to help move their projects toward commercialization.

"These projects have the potential to solve some challenging issues, and we look forward to seeing how the funding and mentorship help them complete their milestones in the coming year," says Denise Graves, university relations director at the Michigan Economic Development Corporation.

Finding a Need

Asher and Goberville's business journey began in early 2019 when �������� was tapped by the Michigan Department of Transportation to develop and pilot two autonomous electric shuttles on campus for students with disabilities.

"I think Nick was initially shocked. At the time, he was working on some of the details with the autonomous system and came to find out that even if there are a little bit of weather issues—something as minor as slight rain—basically the whole pilot would completely shut down," Asher says. "At the same time, there was a lot of research going on over at the American Center for Mobility talking about the need for weather research for autonomous vehicles. And I think Nick and I were both very inspired to try and pursue this."

The wheels began turning, and the team started laying the groundwork for Revision Autonomy. They secured a grant through ��������'s technology development fund and also enrolled in the Introduction to Customer Discovery—ICD—training program, which offers guidance to entrepreneurs as they begin to explore target markets and customers.

“From the ICD program, I began to gather the business perspective and realized it doesn’t matter how awesome a solution you develop, it only matters if others are interested in buying that solution,” says Goberville. “I’ve also learned to communicate with experts in the industry like program managers, engineers, directors and CEOs at some amazing top companies like NVIDIA, Honda, Toyota and some great automotive startups.”

Through interviews with 30 people in the industry, the team was able to learn more about the needs of potential customers and gaps in the current market. Now, Asher and Goberville are taking part in the, which takes the discovery process to the next level.

Navigating the Unknown

In terms of the natural progression of autonomous technology, Asher believes there's a reason weather conditions haven't been a major focus of researchers so far.

"There are so many things to do for automating aspects of driving around the city, on the highway, etc., that it makes the most sense to most companies to start with the easiest possible weather conditions and get a system working from there before moving to different weather conditions," he says. Because of this, a majority of testing currently underway is happening in places like California and Arizona where clear skies and sunshine are plentiful.

It's created a clear lane for innovators like Asher and Goberville to shift their research to other crucial safety areas that aren't getting as much attention. Plus, it gives them a chance to bring their own personal experience to the table.

"As someone who's lived in a climate that has snow every winter my whole life, I like winter sports," says Asher. "But, being out and active during the winter is always a little bit risky. If there's a lot of snow on the ground, there's ice, and it's not necessarily fun (to navigate). Having a system to help ensure safety when traveling in some of these conditions, I think, is a really big deal."

�������� Michigan's dynamic weather patterns make Kalamazoo a prime location for testing and developing technology able to navigate everything from rain and fog to snow and ice. The proximity to several automotive engineering research and design companies also provides an ideal launching point for Revision Autonomy.

The shift to distance education amid the COVID-19 pandemic hasn't slowed their progress. Goberville is able to collect data from home in suburban Chicago using a vehicle the team acquired that's been outfitted with a number of sensors. He's even able to simulate snowy conditions during the summer months with the help of unmarked gravel roads.

Growing Talent at ��������

Asher and Goberville hope to tap into the unique talent base at �������� if their business continues to expand. The University was among the first in the world to offer autonomous vehicle classes to undergraduate students.

"The program that we have here at �������� is very unique, because we are giving students hands-on experience to learn the fundamental engineering concepts for every part of the autonomous vehicle system," Asher says. "Every student here learns every system."

Goberville, who also earned his bachelor's degree from �������� in 2019, has seen the program grow over the years. Initially interested in robotics, he found his passion for autonomous vehicle technology in Asher's lab. 

“I honestly don’t believe I would have this amount of success and experience if I was at another university with a different professor. I’ve been able to not just see, but assist in the development, logistics, engineering, management and process of the end-to-end development of autonomous shuttles,” says Goberville, who, with Asher’s guidance, was also published in a journal and worked on projects related to artificial intelligence and machine learning, computer vision and vehicle controls, among others. “Dr. Asher’s perspective of learning by doing is what really solidifies the knowledge for me.”

That knowledge has also solidified a potential job opportunity in this new business venture—well before he expects to graduate in 2022.

"This case is a little unique because typically students only do research while they're getting a Ph.D., and then they get funding through MTRAC and these other mechanisms as part of a postdoc," Asher says. "But Nick is the type of student who works extremely hard and he's really intelligent, and he's able to balance both in a way that I have not seen before.

For more �������� news, arts and events, visit �������� News online.