LargeClouds2BIM - Efficient workflow transforming large 3D point clouds to Building Information Models with user-assisted automatization

On average, a large construction project exceeds the original cost framework by 80%, and takes 20% longer to complete than planned. Digital technologies promise to overcome this unsatisfactory situation. In fact, 3D scanners and building information models (BIMs) have now become common on large construction sites and in building maintenance. With the rapid progress of 3D measurement technology, 3D scanners are delivering more accurate and high-resolution 3D point clouds than ever before. The automated transformation of these datasets – each comprising several hundred million 3D points – to the virtual BIM reality is still an unsolved problem.

The main challenges include time-consuming pre-processing steps, the lack of tools for merging 3D recordings, various time and cost-intensive and error-prone manual steps, as well as lack of user-friendly interfaces.

With an interdisciplinary team and the active involvement of future users, LargeClouds2BIM develops algorithms and data structures for the progressive, real-time visualisation of massive point clouds, allowing the users to work on such data without significant pre-processing time.

The robust and precise registration of multiple 3D point clouds with similarity transformations makes data capture more independent of specific 3D scanner technologies. The project combines minimal user input with powerful optimisation methods drawn from geometry processing, following a novel approach to user-guided reconstruction of BIM objects from 3D point clouds. And finally, LargeClouds2BIM also explores flexible and active interfaces to open and proprietary BIM ecosystems.

At the system level, the project aims to investigate and provide (lab-scale) proof of concept of the proposed innovative workflow for the transformation of huge 3D point clouds to BIMs. At the component level, the interdisciplinary project team is developing theoretical concepts and prototypical implementations of algorithms and data structures for progressive real-time visualisation, similarity transformation registration, largely automated reconstruction of BIM objects, as well as user-friendly, active interfaces to open and proprietary BIM ecosystems.

The final evaluation of the entire workflow will investigate to what extent the higher degree of automation, flexibility and accuracy can achieve the expected cost and time savings of an estimated 20-30%.

Consortium lead

Rechenraum GmbH

Other consortium partners

• Institute for Interdisciplinary Building Process Management - Integrated Planning and Industrial Building (TU Wien)
• Institute of Visual Computing and Human-Centered Technology (TU Wien)
• LN2 Baumanagement GmbH
• Point of Measure GmbH

Rechenraum GmbH
Dr. Simon Flöry
Gartengasse 21/3
A-1050 Vienna