Artificial Intelligence for Road Safety Analytics: Crash Risk, Near-Miss Detection, and Proactive Interventions

Despite decades of engineering, enforcement, and education-based safety initiatives, road traffic crashes have remained a great burden in the global society economically and socially. The conventional crash-based safety evaluation procedures although a good form of safety evaluation are tended to be reactive in nature and limited by data constraints like under reporting, temporal constraints and infrequency of serious crashes. The latest achievements in artificial intelligence, machine learning, and computer vision provide new opportunities to change the current paradigm of road safety management and make it a more proactive and predictive one. This book gives a systematic and comprehensive way of looking at the use of AI in the analysis of the road safety. The emphasis will be put on modelling of crash risks, identification of near-misses based on surrogate safety measures, and the use of data to support the proactive safety measures. An emphasis on explainable and reliable artificial intelligence models capable of promoting the practical implementation by transportation agencies and policymakers is emphasized. The book aims at the researchers, students undertaking postgraduate studies as well as those who are employed in the field of transportation engineering, traffic safety, and the intelligent transportation system. Blending the theoretical basis and the recent research results, this study will help fill in the gap between sophisticated analytics and practical applications of road safety.