|
ORANGEBURG, S.C. – South Carolina State University’s Dr. Judith Mwakalonge and Dr. Saidi Siuhi are among a team of researchers recently awarded a $1.3 million federal grant to study commercial motor vehicle (CMV) crashes.
Funded by the Federal Motor Carrier Safety Administration (FMCSA), the project aims to use artificial intelligence to reduce CMV crashes in highway work zones via alerts through wireless technology.
“This project will play a key role in expanding SC State University transportation research capabilities by applying artificial intelligence in addressing commercial vehicle safety in work zones,” said Mwakalonge, a professor of transportation engineering and the project’s principal investigator. “Our transportation students involved in this will be uniquely empowered with current technologies to solve transportation problems.
“We look forward to collaborating with Benedict and Clemson on this timely work,” she said.
Siuhi, an assistant professor of civil engineering, serves as the project’s co-principal investigator. Drs. Gurcan Comert at Benedict College and Mashrur Chowdhury at Clemson University also are on the team.
"It is exciting to be a part of the South Carolina State University team in the FMCSA grant that will improve safety on U.S. roads through our proposed artificial intelligence-based real-time work-zone safety assessment,” Chowdhury said.
“It has always been rewarding for us to collaborate with Dr. Judith Mwakalonge and her team at South Carolina State University on new frontiers of transportation research. We are fortunate to expand our collaboration through this FMCSA grant,” Chowdhury said. “We can’t wait to start this new collaboration with South Carolina State University and Benedict College on this project."
While the study initially will utilize data from the South Carolina’s Department of Public Safety and Department of Transportation and the work will take place along highways in South Carolina, officials expect the effort to result in a prototype used nationally.
Most commercial vehicles are large, leading to challenges associated with operations that contribute to crashes, injuries, and fatalities on U.S. highways. National Highway Traffic Safety Administration (NHTSA) Fatality Analysis Reporting System (FARS) data show that commercial vehicles’ involvement in fatal work zone crashes is increasing:
- In the U.S., a work zone fatality takes place every 15 hours, while injuries from work zone accidents occur every 16 min.
-
In 2019, among the 115,000 work-zone-related crashes, CMVs were involved in 22,000 crashes, among which 4000 were injury crashes, and 247 were fatal crashes, with 288 fatalities.
Collecting information from traffic cameras and roadside units, the grant team will develop a safety application for commercial vehicles:
- A vision-based deep learning model will detect vehicles under different weather and lighting conditions.
- The team’s algorithm will generate safety alerts during risk of a crash.
- The team will use video data from multiple traffic cameras at safety-critical road sections/intersections to detect CMVs, generate safety messages, and provide CMV-related safety alerts to approaching vehicles.
- On a section of the road, connected vehicles wirelessly connected with other vehicles, people, and infrastructure can receive wireless alerts through an in-vehicle screen or cell phone.
- Non-connected vehicles will receive the alert via dynamic message signs or cell phones.
| 
