Recently, a paper titled “Vision-Based Collision Prevention on Construction Sites: Integrating Trajectory Prediction and Uncertainty Modeling,” co-authored by Dr. WANG Zeli from the Department of Management Science and Engineering at the School, Dr. YAN Xuzhong from Zhejiang University of Technology, Professor LIU Qinyuan from Tongji University, and Professor LI Heng from The Hong Kong Polytechnic University, was published in the top engineering management journal Journal of Construction Engineering and Management.

      Addressing the collision risks caused by mixed “pedestrian–vehicle” traffic at construction sites, this study proposes an active collision prevention system that integrates computer vision with advanced trajectory prediction. By introducing state-of-the-art visual perception techniques and a Transformer-based spatiotemporal trajectory prediction model, the paper presents an innovative method for real-time and accurate trajectory forecasting of both pedestrians and vehicles. Building on this, the research designs a dynamic risk-zone partitioning approach that combines motion-state analysis with uncertainty evaluation, enabling the identification of potential collision-risk areas—particularly in vehicle-turning scenarios where traditional methods are prone to misjudgment. Experimental results demonstrate that the proposed system achieves excellent performance in target detection, trajectory prediction, and risk identification, and can proactively recognize potential hazards in complex construction environments, thereby significantly enhancing the proactive safety management capabilities of construction sites.

     Dr. WANG Zeli received his PhD from The Hong Kong Polytechnic University and joined the School in 2021. He was selected for the Shanghai Pujiang Talent Program in 2022 and named a Shanghai Chenguang Scholar in 2025. His long-term research interests include computer vision, intelligent construction, and smart management. He has presided over several national and ministerial-level research projects, including a Young Scientists Fund project of the National Natural Science Foundation of China, and his work has been published in international authoritative journals such as Automation in Construction, Pattern Recognition, and the Journal of Construction Engineering and Management.

     Paper Information: Wang, Z., Yan, X., Liu, Q., Ma, X., & Li, H. (2026). Vision-Based Collision Prevention on Construction Sites: Integrating Trajectory Prediction and Uncertainty Modeling. Journal of Construction Engineering and Management, 152(1), 04025214.