Sistema de escáner láser mediante dispositivo slam sr10

Abstract

This internship report presents the development, standardization, and validation of a workflow for 3D BIM modeling based on point clouds generated with the SLAM SR10 mobile laser scanner, aimed at reducing processing times by 30% and improving geometric accuracy in the digitalization of industrial infrastructure. Through a technical diagnosis, deficiencies in data capture and management were identified—such as low density, incomplete coverage, noise presence, and trajectory misalignment—allowing for the structuring of an optimized methodology. This methodology ranges from field planning and capture with controlled parameters for speed, overlap, and session duration, to advanced processing using SLAM-PPK trajectory adjustment in CoPre2, conversion and cleanup in Autodesk Recap, and integration for modeling in Autodesk Revit. Additionally, the potential of artificial intelligence for automating the classification and semantic structuring of construction elements was evaluated. The results, validated in real-world industrial projects, demonstrate significant improvements in operational efficiency, geometric consistency, and quality control, achieving root mean square errors (RMSE) of less than 5 cm and establishing a replicable technical guide for the generation of high-fidelity three-dimensional representations in BIM environments.