Sight Distance Reliability Analysis of LiDAR-Based Automated Vehicles on Road Horizontal Curves under Adverse Weather Conditions

The sight distance reliability of LiDAR-based automated vehicles (LAVs) in complex road environments is critical for their deployment. Road geometry and weather conditions are two key factors affecting the perception capabilities of LAVs. However, current studies rarely analyzed the combined effects of these two factors on sight distance performance. This study investigates LAVs’ sight distance reliability on curved roads in adverse weather, considering various design speeds, curve radii, and scenarios including clear weather, rain, and fog. Available sight distances (ASDs) are extracted using defined LiDAR point cloud thresholds to develop sight distance reliability functions. Moreover, the Monte Carlo simulation quantifies sight distance failure risks associated with different levels of LAVs in varied operational contexts. The results unveil a significant impact of weather conditions on ASDs, highlighting that decreased visibility and increased rainfall adversely affect ASD, with a notable 56.64% probability of sight distance failure under certain conditions. Additionally, the study finds that shorter perception-reaction times can mitigate sight distance risks when LAVs navigate on curved roads, whereas higher speeds exacerbate these risks. Furthermore, the study reveals that lower automation levels struggle to maintain adequate sight distances on existing curved roads under adverse weather conditions. These insights remind road managers to determine appropriate speed limits for LAVs on curved roads, enhancing operational safety from a sight distance perspective.