EVALUATING 3-D SIGHT DISTANCE AT URBAN INTERSECTIONS USING A LIDAR-BASED MODEL AND CONSIDERING MULTIPLE USERS

The provision of sufficient available sight distance along the road reduces the potential for conflicts among its users. Consequently, sight distance estimations are significant components of road safety analyses. Evaluating sight distances on existing roads, when not performed in situ, requires up-to-date representations of the roads' geometric properties and roadside elements. For some time now, LiDAR instruments have been used to gather geospatial data needed to generate updated models of the roadways and roadside elements. Digital terrain models (DIM) have been widely utilized to perform sigh distance estimations, despite the fact that these models only account for the road surface and bare terrain. On the other hand, Digital Surface Models (DSM) which provide further information of features, namely vegetation, traffic signs, buildings and other manmade elements, yield inaccurate results when some elements protrude or project over roadway areas like cantilevered signs, tree branches and in the presence of complex above ground features. The authors have previously proposed a methodology to overcome this inaccurate representation of road assets and surrounding elements utilizing Geographic Information System (GIS) tools and 3-D geometries on top of the DIM so as to represent possible visual obstructions. In this analysis, the available sight distance of an urban intersection located in Madrid (Spain), is evaluated considering its prevalent motorized vehicles. Virtual trajectories of pedestrians and cyclists were modelled in order to assess the impact of road surrounding objects in their available sight distance. Furthermore, the separate consideration of 3D objects allows the possibility of assisting decision making processes concerning the location of urban furniture while assuring proper visibility.