LiDAR-based lane marking encoding using reflectivity variation

This study investigates a method to enhance the functionality of conventional road marking materials by enabling them to carry encoded information. A technique is proposed for encoding thermoplastic lane markings by modulating their surface retroreflectivity. This is achieved by controlling the surface distribution of glass beads to create a binary signal readable by LiDAR sensors. Laboratory experiments determined that a glass bead surface distribution between 1.6% and 2.0% yields a stable and distinguishable high-reflectivity signal. This finding allowed for the design of a minimum encoding unit length of 2 cm, balancing information density with material performance. The reliability of this infrastructure-based encoding method was successfully validated on a full-scale prototype using a custom information parsing algorithm. This research demonstrates the feasibility of enhancing the functionality of conventional pavement materials, offering a low-cost and non-disruptive design approach for future intelligent road infrastructure.