A novel skateboarder-related near-crash identification method with roadside LiDAR data

Abstract

Skateboarding is being an emerging travel model, especially for young travelers. The conflict between skateboarders and the other road users has raised safety concerns for traffic engineers. Safety evaluation about skateboarder-related conflicts has not been well performed due to the low skateboarder-related crashes and the limited historical crash data. Near-crashes have been considered as surrogate data for skateboard-related safety evaluation. This paper developed a procedure to extract skateboarder-associated near-crashes automatically with the roadside Light Detection and Ranging (LiDAR). A new indicator: distance-deceleration-time profile (DDTP) which combined time, space, and deceleration information was introduced for skateboarder-pedestrian near-crash identification. The DDTP was developed for the roadside LiDAR data specially. The case studies showed that the proposed method can extract skateboarder-pedestrian safety-critical events with high accuracy. The proposed method can be also used for skateboarder-vehicle and skateboarder-bicycle near-crash identification.