Stojanovic, Vladeta; Trapp, Matthias; Richter, Rico; Hagedorn, Benjamin; Döllner, Jürgen
to be published
Advances versus adaptation of Industry 4.0 practices in Facility Management (FM) have created usage demand for up-to-date digitized building assets. The use of Building Information Modelling (BIM) for FM in the Operation and Maintenance (O&M) stages of the building lifecycle is intended to bridge the gap between operations and digital data, but lacks the functionality of assessing and forecasting the state of the built environment in real-time. To accommodate this, BIM data needs to be constantly updated with the current state of the built environment. However, generation of as-is BIM data for a digital representation of a building is a labor intensive process. While some software applications offer a degree of automation for the generation of as-is BIM data, they can be impractical to use for routinely updating digital FM documentation. Current approaches for capturing the built environment using remote sensing and photometry-based methods allow for the creation of 3D point clouds that can be used as basis data for a Digital Twin (DT), along with existing BIM and FM documentation. 3D point clouds themselves do not contain any semantics or specific information about the building components they represent physically, but using machine learning methods they can be enhanced with semantics that would allow for reconstruction of as-is BIM and basis DT data. This paper presents current research and development progress of a service-oriented platform for generation of semantically rich 3D point cloud representations of indoor environments. A specific focus is placed on the reconstruction and visualization of the captured state of the built environment for increasing FM stakeholder engagement and facilitating collaboration. The preliminary results of a prototypical web-based application demonstrate the feasibility of such a platform for FM using a service-oriented paradigm.