The goal: Reducing the time spent between data acquisition and Building Information Model through new automation methodologies.

The project BIMiB: Contrary to newly built structural environments, the data on existing buildings intended for modification is rarely extensive enough for direct economic use in modeling. For this reason, a system is being developed which reconstructs a digital Building Information Model (BIM) from 3D laserscan point clouds and digitized drawings of the structure. New approaches to ontology-based semantic knowledge representations in combination with machine learning and pattern recognition are being researched for their combined potential to increase automatic detection rates. Through these efforts, structure repurposing and revitalization becomes more economically viable for increased building life-cycles.

The parties involved: BIMiB was launched in May 2019 by Prof. Dr. rer. nat. Dominic Becking and Prof. Dr.-Ing. Uwe Weitkemper at the Institute for the Intelligent Building. The project coordinators from different fields of expertise represent computer science and structural engineering, respectively. In addition, an interdisciplinary team of doctoral staff and student assistants supports the work on BIMiB at InfinteG. Funding was provided by the German Federal Ministry of Education and Research in cooperation with Hochtief ViCon, Essen, and Pape Architekten, Herford. Both partner companies advise the research team and provide an end user perspective on developed concepts and methods.

Our request: Machine learning needs data. To anchor the development of possible solutions in a sustained metric towards sufficient generalization capabilities, we are constantly in search of existing constructions of which digitized data can be introduced to the project. If our proposed methods can prove their robustness in realistic reconstruction scenarios, their added value may be provided to the building industry more readily. We already collaborate with our partner companies at the boundary between science and economy to incorporate the ideas of innovative catalysts. In parallel to this joint effort, we are always open to form new contacts to further interlink our work with the intended user base. Do not hesitate to get in touch with us if you have an interest in our work.

The requirements: This project needs data of existing buildings with a reinforced concrete solid or skeleton construction style. The following elements should ideally be present: reinforced concrete slabs, beams, columns, (non-)load bearing walls and staircases. We utilize data in form of as-built plans, 3D laser scan point clouds and a Building Information Model of the structure. Our prioritized sources are 2D plans of adequate quality, but the retrospective laser scanning of the structure must be possible. Plans should be accessible for digitization, meaning no excessive staining or markings preventing a feasible analysis. All plans can be provided as PDF files or in raw paper scans, must be complete and are ideally supplemented by a positioning plan. A fully enriched Building Information Model is optional.

For a data contribution, we have essential (), optional () and prohibitive () criteria:

  • As-built drawings
  • Accessibility for Laserscanning
  • Data not protected by IP rights
  • Structural calculation
  • Building Information Model existing
  • Point clouds existing
  • 2D plans incomplete
  • Special buildings
  • Structural system visually inaccessible

Should you consider to support this project with a data contribution or have any other interest in our work, we are happy to hear from you!

Support science.                                   Influence developments.                                   Co-create progress.

Contact:

Ms. Silke Dierssen

Research Fellow

→silke.dierssen@fh-bielefeld.de