Proceedings of the 2022 European Conference on Computing in Construction

     

ScaleBIM: Introducing a scalable modular framework to transfer point clouds into semantically rich building information models

Fabian Kaufmann, Christian Glock, Thomas Tschickardt
Technical University of Kaiserslautern, Germany

DOI: 10.35490/EC3.2022.194
Abstract: Current approaches for the automated acquisition of Building Information Models (BIMs) of existing buildings are limited to only few elements of the building. Furthermore, most work is also related to the reconstruction of indoor elements, and not dedicated to structural elements (e. g. walls, slabs, beams, columns, openings). In this work, a framework for scan to BIM automation is proposed that focusses on structural elements and has a modular structure that allows the adaptation to other types of elements or types of structures. To achieve this, advanced methods for segmentation and data exploitation are being applied.
Keywords: scan to BIM, machine learning, point cloud, reconstruction, IFC
Paper:
EC32022_194

Presentation Video: