Developing a Methodology to Estimate the Retroreflectivity of Longitudinal Pavement Markings using LiDAR

Longitudinal pavement markings significantly affect traffic safety, particularly in adverse weather and nighttime conditions when crashes and fatalities are often overrepresented. Given the wide range of factors affecting the performance of pavement markings, periodic monitoring is needed to ensure their integrity and adequate retroreflectivity levels. Typical monitoring methods include individual readings from manual retroreflectometers and, more recently, mobile setups where much larger segments can be covered in shorter periods. However, mobile setups require specialized equipment, calibration, and significant economic resources. This research uses LiDAR data collected as part of asset management efforts to help assess pavement marking retroreflectivity, reducing reliance on special-purpose equipment, mainly for maintenance-related decisions. Identifying and isolating the pavement marking from the LiDAR point cloud, filtering, and modeling are part of a proposed exploratory process to evaluate the associations between field-measured retroreflectivity and a combination of intensity from LiDAR readings, RGB data, and marking material. Freeway data covering over 300 miles of markings along freeways are used to analyze the initial potential of LiDAR in assessing the retroreflectivity of pavement markings. Results are encouraging and show that LiDAR data can produce reasonable associations to retroreflectivity levels. However, significant questions remain open in terms of the transferability of results, modeling other variables such as weather and traffic exposure, and the inherent differences that may arise when evaluating different treatment materials and markings at different points within the roadway (i.e., at varying distances and angles from the LiDAR sensor).