Leveraging Connectivity to Enhance Safety

Our research explores data-driven methods to understand and improve driving behavior and roadway safety. We develop predictive models that identify risky or aggressive driving, design algorithms for classifying crash severity, and create real-time detection systems that support proactive safety interventions. By combining connected vehicle data, machine learning, and advanced statistical modeling, we provide insights that help transportation agencies reduce crashes and enhance overall traffic safety.

Illustration of curbspace management

Technical approach

Pattern recognition, machine learning, deep learning.

Example projects

  • Safety Analysis at Intersections Using Vehicle Trajectories
  • Safety Application Using Connected Vehicle Trajectories: A Demo

Key publications