1666994 - Design Speed Evaluation on Horizontal and Vertical Curves Using LiDAR and Imagery-based Point Clouds

The process of evaluating roadway geometry for potential safety problems requires precise measurement of various geometric parameters. This study evaluated the use of mobile LiDAR scanning (MLS) and unmanned aerial vehicle (UAV) imagery-based point cloud data for estimating design speed on horizontal curves as well as sight distance and design speed on vertical curves of constructed roadways. The analyses were performed using data collected at two sites located in Anderson, and Spartanburg County, South Carolina. Crash data for the years 2018-2022 from both counties showed that nearly half of all fatal and serious injury crashes occurred within horizontal curve locations of divided and undivided roadways. Results from paired t-test statistical analyses using a 95% confidence level indicated that mobile LiDAR and UAV photogrammetry systems provide horizontal curvature data at sufficient accuracy for estimating curve design speeds. The proposed methodology can be used to identify locations where the posted speed limit/advisory speed is higher than the design speed along horizontal curves to facilitate the implementation of corrective measures on existing roadway networks. Additionally, vertical alignment data were extracted using terrain models generated from point clouds for sight distance and design speed estimation on crest and sag vertical curves. Extracted longitudinal grades were compared based on a desired target accuracy of ± 0.5% specified by SHRP2. The average deviations between field survey measurements and longitudinal grade measurements extracted from LiDAR and imagery-based point clouds were less than the target accuracy value of ± 0.5% over the same length at a 95% confidence level. Similarly, statistical analyses using paired t-tests showed that sight distances calculated using terrain models from point clouds could be used to obtain reliable estimates of design speed on vertical curves. The proposed approach offers advantages over extracting information from design drawings that may be unavailable, outdated, or inconsistent with the as-built roadway.