3D Virtual Sight Distance Analysis Using Lidar Data
Olsen, Michael J.
Hurwitz, David S.
Kashani, Alireza G.
MetadataShow full item record
This research project investigated advanced safety analysis methodologies for drivers’ sight distance based on high resolution data acquired using lidar (light detection and ranging) technology. Sight distance analyses require careful and detailed field measurements to facilitate proper engineering decision making regarding the removal of obstructions, establishment of regulatory and advisory speed limits, and the location of new access points, among numerous other examples. However, conventional field measurements present safety concerns because they require personnel to be in or adjacent to traffic lanes. They can also be time consuming, costly, and labor intensive. Furthermore, the predominantly two-dimensional (2D) methods involve simplifying assumptions such as a “standard” vehicle heights and lengths without considering the wide range of vehicles and drivers present on the road. Recently, departments of transportation have begun to acquire mobile lidar data for their roadway assets. As an example, Oregon DOT has recently completed scan surveys of all state owned and maintained highways and updates of high priority areas annually. These data provide a rich, threedimensional (3D) environment that enables one to virtually visit a site at any frequency and efficiently evaluate sight distances from the safety of the office. This research presents a systematic processing and analysis workflow for virtually evaluating available sight distances by using lidar data sets named SiDAL (Sight Distance Analysis using Lidar). This approach enables one to repeatedly analyze the same scene whle considering a variety of vehicle types as well as multimodal forms of transportation (e.g., bikes, pedestrians). The sensitivity of this technique to modeling resolution was analyzed by using a case study of an intersection with restricted visibility. The results showed the ability to capture significantly more detail about visibility constraints in comparison to conventional measurements.