Assessing the Feasibility of Utilizing UAS-Based Point Cloud in Pavement Smoothness/Roughness Measurement

Project Details









Erzhuo Che


University Transportation Centers Program, USDOT


Data collection, Drones, Pavements, Roughness, Smoothness

Project description

This proposed effort will assess the feasibility of utilizing unmanned aerial system (UAS)-based point clouds in measuring pavement roughness, and this will be accomplished through the following tasks: (1) Establish a framework to process and calculate International Roughness Index (IRI) from UAS data. The proposed framework will take UAS-based point clouds as input, remove noise, extract pavement, create a terrain model, and generate profiles to extract roughness information. All the methods and approaches will be tailored targeting a more accurate and reliable IRI assessment. (2) Perform rigorous accuracy assessment for UAS data. This project will primarily use the UAS data that was collected in the Oregon Coastal area where terrestrial lidar data is also collected and can be served as ground truth to compare against: (a) Point-based: compare terrestrial lidar point cloud on a per-point basis; (b) Model-based: compare terrain models and IRI assessment results generated from terrestrial lidar point cloud and UAS data; and (c) Sensitivity analysis: test the sensitivity of different parameters and approaches used in noise removal, modeling, and roughness assessment. (3) Provide recommendations for UAS data acquisition and processing procedures for the purpose of assessing pavement roughness. For example, regarding to data collection, the selection of ground sampling distance (GSD) can be very important because it directly affects the resolution and precision of the model. In post-processing, there are many parameters and settings that can substantially impact the point cloud and model generated from UAS data. This task will optimize these acquisition and processing parameters and provide recommendations