Making traffic safer: FAMU-FSU College of Engineering team uses artificial intelligence to improve intersection safety

A team of researchers from the FAMU-FSU College of Engineering is exploring how existing technology and new algorithms can prevent potential crashes at intersections and pave the way for smarter, safer roads. (Adobe Stock)
A team of researchers from the FAMU-FSU College of Engineering is exploring how existing technology and new algorithms can prevent potential crashes at intersections and pave the way for smarter, safer roads. (Adobe Stock)

Florida State University researchers are developing innovative technology to make intersections safer by aiming to reduce crashes and save lives.

A team of researchers from the FAMU-FSU College of Engineering is exploring how existing technology and new algorithms can prevent potential crashes at intersections and pave the way for smarter, safer roads.

“Millions of vehicles move through intersections every day. Our goal is to make that process safer,” said Olugbenga Moses Anubi, an associate professor of electrical and computer engineering and member of the research team.

Olugbenga Moses Anubi, a FAMU-FSU College of Engineering associate professor in the Department of Electrical and Computer Engineering
Olugbenga Moses Anubi, a FAMU-FSU College of Engineering associate professor in the Department of Electrical and Computer Engineering

About 25% of traffic deaths and half of all traffic injuries in the United States are attributed to intersections, according to the Federal Highway Administration.

HOW IT WORKS:
The team’s solution to the problem of intersection danger is called PREDISS, or the Predictive Intersection Safety System.

PREDISS combines data-driven analytics with physics-based methods to improve road safety. The system follows four key phases: identify, predict, anticipate, and warn/mitigate.

With initial funding awarded in 2024, the researchers used existing collision data and autonomous vehicle systems to develop an algorithm capable of predicting vehicle trajectories and identifying potential conflicts. PREDISS uses sensors, cameras, wireless technology and artificial intelligence to track vehicles and pedestrians approaching an intersection, allowing it to predict their trajectories and identify potential conflicts in real-time.

Now they are moving to the next phase — anticipation.

In collaboration with the City of Tallahassee’s Regional Transportation Management Center, the team will install PREDISS at a live traffic intersection to monitor real-time conditions. This setup will generate scenarios for testing algorithms and help researchers refine predictive models and develop proactive safety measures.

The final phases — warning and mitigation — will depend on data from the live intersection testing. Future implementations could include adaptive warning alarms, additional red-light enforcement and even automated intervention in autonomous vehicles to prevent collisions.

Ren Moses, a FAMU-FSU College of Engineering professor in the Department of Civil and Environmental Engineering
Ren Moses, a FAMU-FSU College of Engineering professor in the Department of Civil and Environmental Engineering

The hardware PREDISS uses has mostly been developed, which will help smooth the process for deploying and testing the system.

“We’re taking existing tools and applying them in creative ways to help solve this problem,” said project member Ren Moses, a FAMU-FSU College of Engineering professor in the Department of Civil and Environmental Engineering.

WHO’S INVOLVED:
Anubi and Moses are working with Paul Bupe Jr., a research faculty member dedicated to PREDISS, and Joshua Hollingsworth, manager of traffic operations for the City of Tallahassee.

Working with the city allows the researchers to collect data from a real-world scenario, helping to validate their algorithms and ensure that their system can handle the many unforeseen challenges presented by a busy intersection.

“Our cooperation with the city is a great example of how researchers can partner with outside collaborators and stakeholders to meet shared goals,” Anubi said. “This testing really accelerates the process of making PREDISS into a working system that can be used anywhere to improve traffic safety.”

This research is the second stage in the U.S. Department of Transportation’s Intersection Safety Challenge. After winning Stage 1A of the 2024 U.S. Department of Transportation’s Intersection Safety Challenge, the team has now won the 2025 Stage 1B tier II prize. These achievements have brought in $266,666 in funding to advance their project, which integrates AI-driven technology with physics to enhance road safety.

“We are grateful to be in Florida, where the Florida Department of Transportation (FDOT), invests heavily in smart traffic management systems, such as traffic cameras, real-time traffic data and intelligent transportation systems to reduce congestion and improve safety,” Anubi said. “We hope to partner with FDOT in making our roadways, intersections and traffic systems as safe as possible.”

WHY IT MATTERS:
Fatal crashes in intersections take thousands of lives every year — more than 12,000 in 2022, the latest year with data available on the Federal Highway Administration website.

Engineering has a major role to play in mitigating those crashes. Changing the design of intersections is one potential alternative, but that option is not possible for every site, and even in newly engineered intersections, other technologies can continue improving safety.

“Traveling our roadways should be as stress-free and safe as possible,” Anubi said. “The opportunity to help improve safety on roads where we live and where our family and friends travel nearly every day emphasizes the importance of this project.”

An image showing how PREDISS, or the Predictive Intersection Safety System, tracks vehicles at intersections. (Courtesy of Olugbenga Moses Anubi)
An image showing how PREDISS, or the Predictive Intersection Safety System, tracks vehicles at intersections. (Courtesy of Olugbenga Moses Anubi)