(October 23) Blending AI/ML with Traditional Data Science: The Case of Highway Safety Research

Artificial intelligence/machine learning (AI/ML) approaches to data analytics have gained considerable traction in highway safety research. However, the behavioral elements of the highway safety problem are fundamental, with driver behavior and the associated behavioral responses to potential safety interventions presenting a formidable challenge. Previous highway safety research, using traditional data science based primarily on statistical and econometric approaches, has struggled to capture the human-behavior element of the highway safety problem, even while drawing on behavioral science to address underlying issues relating to causality and associated concepts such as endogeneity, unobserved heterogeneity, sample selectivity, and temporal dynamics. Conversely, contemporary AI/ML applications have typically emphasized prediction while overlooking the behavioral mechanisms that determine safety outcomes, a disconnect that has limited both interpretability and predictive performance. This presentation discusses the behavioral complexities involved in the highway safety problem, the strengths and weaknesses of traditional and data-driven analytics and then argues that integrating behavioral constructs and traditional data‑science principles with contemporary AI/ML methods can substantially enhance both inference and prediction in highway safety analytics. Recent examples of this integration are presented, illustrating how such an approach can provide new insights and open new avenues for highway safety research.

Fred Mannering, PhD
Executive Director, Center for Urban Transportation Research

Fred Mannering is a Professor of Civil and Environmental Engineering and Executive Director of the Center for Urban Transportation Research (CUTR) at the University of South Florida (USF). He previously held academic positions as the Charles Pankow Professor and Head of Civil Engineering at Purdue University, Professor and Chair of Civil and Environmental Engineering at the University of Washington, and Assistant Professor at the Pennsylvania State University. He received his undergraduate degree from the University of Saskatchewan, masters from Purdue University, and PhD from the Massachusetts Institute of Technology. His research interests are in the application of statistical and econometric methods to study highway safety, transportation economics, travel behavior and a variety of engineering-related problems. He has published over 170 journal articles and two widely-adopted text books: Principles of highway engineering and traffic analysis (Amazon) (the seven editions of which have been used to educate more than 150,000 students at universities in the United States and around the world ) and Statistical and econometric methods for transportation data analysis (Amazon) He is the founding Editor, current Chief Advisory Editor, and previous Editor-in-Chief (2013 to 2024) of the Elsevier Science journal Analytic Methods in Accident Research. He also was previously Editor-in-Chief (2004-2012) of the Elsevier Science journal Transportation Research Part B – Methodological.