Diving into safe-speed recommendation and data-driven solutions for safer and smarter mobility

For the upcoming Transportation Research Board Annual Meeting 2026 in Washington D.C., Zhang will present her recent work, “Quantile-physics hybrid framework for safe-speed recommendation under diverse weather conditions leveraging connected vehicle and road weather information systems data.” 

This study analyzes over 6.6 million connected vehicle records and more than 10,000 road weather sensor records from the Buffalo-Niagara region to model how adverse weather, road surface conditions and vehicle behavior influence safe driving speeds. 

Zhang developed a scalable big-data pipeline and trained a hybrid machine learning model that generates lower and upper bounds for safe driving speeds under varying conditions such as snow, rain and low visibility. 

The framework offers actionable guidance for both human-driven and autonomous vehicles to enhance roadway safety. Zhang’s earlier publication, “Leveraging high-resolution connected vehicle data for incident detection, management and analysis: An exploratory study,” highlighted the advantages of connected vehicle data compared to traditional fixed-location sensors in traffic incident analysis. Zhang’s ongoing research also explores reinforcement learning for optimizing on-demand flexible transit services. 

Using real-world data to build simulation environments, she aims to improve vehicle dispatching and routing efficiency, ultimately enhancing both system performance and rider experience. 

Zhang is passionate about conducting research that leads to practical, deployable solutions for real-world transportation challenges. Her long-term goal is to advance transportation safety and efficiency through the integration of machine learning, optimization and large-scale data analytics.