UToledo Transportation Systems Research Lab Leading New City, State Projects

August 21, 2020 | News, Research, UToday, Engineering
By Christine Billau



Civil engineers in The University of Toledo Transportation Systems Research Lab are looking for undergraduate and graduate students to join their work on several projects critical to the future of driving on roads in Ohio.

The Ohio Department of Transportation awarded the lab $250,000 to study how to use artificial intelligence, deep learning and computer vision technologies to detect rumble strips on roads, evaluate roadway conditions for potholes, patches and cracks, and automatically count and classify passing vehicles for traffic management.

Pratik Shrestha, who is pursing a master’s degree in civil engineering, works in the Transportation Systems Research Lab.

“These technologies are similar to those employed by the autonomous driving industry for self-driving vehicles to navigate roads, and they are becoming very prevalent in our everyday life,” Dr. Eddie Chou, professor of civil engineering and director of the Transportation Systems Research Lab, said. “Our goal is to help agencies such as ODOT use these new technologies to maintain safer roads, collect roadway and traffic data, and reduce traffic congestions.”

The researchers also are working with the city of Columbus to implement state-of-the-art tools that will help Columbus become a digital, smart city that efficiently utilizes its resources to improve infrastructures and services, reduces traffic crashes and fatalities, and increases equity among all residents. This one-year, $49,500 pilot project started in April.

“We are very excited about the potential impact of these projects and hope to attract new students to participate in the research,” Chou said. “We are very fortunate to receive the external funding to perform this cutting-edge research, especially during this time when many funding agencies are facing enormous budget shortfalls.”

For this research, Chou’s lab acquired an AI/deep-learning machine equipped with two powerful graphical processing units that perform a large number of mathematic computations rapidly through parallel processing.

“This enormous computing power allows artificial intelligence and deep-learning models to be trained using large amount of data, efficient algorithms and fast-computing hardware,” Chou said.

Deep learning is machine learning — the most popular AI field — using “deep” neural network models, which means many layers of neurons within the artificial neural network models. The more layers of neurons, the more computations are involved.

Click to access the login or register cheese