Video-Based Safety Evaluation of Same-Direction Interactions Between Cyclists and Motorized Vehicles

As the safety of cyclists remains a critical concern for sustainable mobility, there is a growing need for practice-ready tools to reliably assess the safety of cycling infrastructures. Video cameras present an accessible and efficient means of data collection; however, converting raw data into relevant safety indicators presents considerable challenges. This study introduces a simple yet flexible tool for deriving such indicators from any set of trajectories mapped onto a planimetric plane. Following data filtering procedures, bicycle-vehicle interactions are categorized as either "overtaking" or "longitudinal", based on whether an overtake occurs or if the motor vehicle remains behind the bicycle. Safety-relevant indicators are then computed for each situation. The application of the proposed framework is illustrated through two real-world case studies situated in the Hannover area, Germany. Video data collected both before and after infrastructural interventions allow for before-after analyses concerning (i) the introduction of a new bicycle lane and (ii) the application of a new paint colour to an existing bicycle lane. These case studies provide valuable insights for discussion and effectively demonstrate the advantages and limitations of the proposed methodology.