Projects

• Developed a SUMO-based traffic simulation framework to evaluate roadside responder safety, modelling car-following, lane-changing, driving decision, and peer influence across varying traffic conditions.
• Analyzed nationwide survey data on roadside responders’ near-miss experiences, reporting practices, and acceptance of automated reporting technology.
• Integrated roadside cameras and LiDAR for field data collection and implemented a real-time framework for vehicle trajectory processing and near-miss detection.

• Developed a travel-demand simulation framework for ride-sharing, integrating a real-time integer-programming model for dynamic ride-sharing assignment.
• Applied the simulation framework to multiple transportation projects in evacuation, traffic incident response, shared commuting, and last-mile delivery.

• Developed a high-performance agent-based traffic simulator for evaluating shared-commuting and carpooling strategies in Birmingham, Alabama.
• Parallelized simulation computation by partitioning urban network zones across processes using Repast4py and METIS.
• Implemented dynamic METIS-based network repartitioning to maintain computational load balance during simulation.

Conducted research on programs supporting violin, recorder, and piano performance using audio processing, time-series analysis, Bayesian network modelling, and Markov modelling.

• Developed a machine learning framework for crash risk prediction using macro- and micro-level factors from multiple traffic safety and vehicle datasets.
• Designed an improved deep-SVDD algorithm to model crash risk at the probability level.