Research

Our group investigates the complex interactions between humans and transportation systems, with particular emphasis on the psychology, behavior, and safety of pedestrians, drivers, and other transport stakeholders. We combine immersive virtual reality (VR), advanced data mining, and artificial intelligence to translate empirical findings into tools for safer, adaptive, human-centered transport systems.

Behavioral human factors and psychology

• Pedestrian behavior under unfamiliar traffic rules — VR studies reveal habitual looking behaviors that increase risk when people cross under driving rules opposite to those in their home country.
• Alcohol impairment and decision-making — Experiments examine how intoxication shifts crossing priorities from safety toward efficiency.
• Human–autonomous vehicle interaction — VR and eye tracking help us study how external human–machine interfaces affect attention, cognitive load, and crossing errors.
• Pedestrian–autonomous truck platoon interaction — Behavioral models reveal how trust, risk perception, and individual tendencies shape gap acceptance.

Traffic accident analysis and data mining

• Bayesian random-parameter and quantile models for heterogeneous and heavy-tailed maritime accident losses.
• Text mining of safety reports to identify recurrent causes and targeted countermeasures.
• Resilience and risk analytics for maritime and multimodal transport systems.

AI-based traffic modeling and prediction

• Inverse reinforcement learning for safety–efficiency trade-offs in behavior.
• Physics-informed neural networks for dynamic traffic assignment and state estimation.
• Risk-aware pedestrian–vehicle interaction and trajectory modeling.
• Interpretable machine learning for transport safety prediction.

Our mission

We bridge transportation engineering, behavioral psychology, data science, and machine learning to support policy-makers, planners, and industry in creating safer mobility systems.