Real-time crash prediction models: State-of-the-art, design pathways and ubiquitous requirements (Accident Analysis and Prevention, 2019)

With the advancements in artificial intelligence, and multiple studies being focused towards developing real-time crash prediction models, the concept of a proactive safety management system has become very close to reality. The linked study conducts an extensive review of the existing real-time crash prediction models, systematically illustrating the various methodologies being used world-wide. It evaluates the universality, design requirements, and associated challenges of various models. The study aims to be a “one stop knowledge source” for future researchers and practitioners for transitioning from the existing real-time crash prediction conceptual models to a real-world operational proactive traffic safety management system.

Download the study.