Raw 3D laser scan point clouds and pendulum test value (PTV) and Traction Watcher One (TWO) friction measurements from 45 asphalt pavement test sections

This dataset contains raw 3D laser scan surface texture point clouds (XYZ text files) and corresponding Pendulum Test Value (PTV) measurements (EN 13036-4) from 45 asphalt pavement test sections (20 measurement points each, 900 locations total). Each test section is divided into two parts (01 and 02), corresponding to measurement points located before and after a reference steel plate. This plate served as a fixed reference marker for reliable identification of measurement locations during continuous friction coefficient measurements. The point cloud for each measurement location consists of 8 scanning strips representing individual laser passes, stored in 8 separate text files. These strips can be aligned and merged to reconstruct the full surface patch. The reference grid used to define the pendulum slider slip path is visible in the raw scans and can be identified and removed during preprocessing. Data are provided in raw, unprocessed form to allow independent analysis with user-defined filtering, descriptor extraction, and modeling workflows. No preprocessing, filtering, or descriptor extraction has been applied. The dataset supports two manuscripts currently under review: "Machine-learning prediction of skid resistance by British Pendulum Tester from 3D pavement surface texture: assessing the relative contributions of micro- and macrotexture" (International Journal of Pavement Engineering) and "Machine-learning based prediction of skid resistance from 3D pavement surface texture: close-range photogrammetry versus high-resolution laser scanning" (Measurement). Friction measurements. Pendulum Test Value (PTV, EN 13036-4) and the Traction Watcher One (TWO) longitudinal friction coefficient µ are provided as companion files, semicolon-delimited in a section–position–point ID format (e.g. "01P0101;0.310"). Documentation. The AI model trained on these data, the full training methodology, performance metrics, and the MATLAB analysis code are documented in a separate Zenodo record: https://doi.org/10.5281/zenodo.21247668 Funding. Funded by the European Union — NextGenerationEU, through the Recovery and Resilience Plan of the Slovak Republic (call 09I03-03-V04, project 09I03-03-V04-00041 — FricPred), administered by the Research Agency (Výskumná agentúra) on behalf of the Government Office of the Slovak Republic.