Safety Assessment of Urban Intersection Sight Distance Using Mobile LiDAR Data

Abstract

This paper proposes an automated framework that utilizes Light Detection and Ranging (LiDAR) point cloud data to map and detect road obstacles that impact drivers’ field of view at urban intersections. The framework facilitates the simulation of a driver’s field of vision to estimate the blockage percentage as they approach an intersection. Furthermore, a collision analysis is conducted to examine the relationship between poor visibility and safety. The visibility assessment was used to determine the blockage percentage as a function of intersection control type. The safety assessment indicated that intersections with limited available sight distances (ASD) exhibited an increased risk of collisions. The research also conducted a sensitivity analysis to understand the impact of the voxel size on the extraction of intersection obstacles from LiDAR datasets. The findings from this research can be used to assess the intersection without the burden of manual intervention. This would effectively support transportation agencies in identifying hazardous intersections with poor visibility and adopt policies to enhance urban intersections’ operation and safety.