Investigating Low-Cost Rutting Assessments Using Low-Density Mobile LiDAR Data

Abstract

Pavement rutting compromises driver safety, yet traditional inspections are slow and existing automated systems require expensive, high-resolution Light Detection and Ranging (LiDAR). This paper introduces a novel tile-based method to automatically quantify rutting using low-cost mobile LiDAR data. The algorithm identifies and measures ruts by modeling pavement deformations using computational geometry and surface fitting. A key contribution is a sensitivity analysis investigating the trade-off between point density and measurement accuracy. Tested on five road sections, the approach accurately measured rut depths as shallow as 3 mm. The analysis confirmed that low-density data, simulating low-cost sensors, can quantify ruts within practical error margins at a point density of 230 points per square meter (ppsm). The runtime for rutting estimation decreases from 83 to 34 seconds per kilometer when using low-density data. This study demonstrates a feasible and efficient solution for network-level rutting assessment, enabling wider and cost-effective adoption of LiDAR technology.