Employment of Semantic Web Technologies for Capturing Comprehensive Parametric Building Models

Building Information Modelling is a well-known acronym in the construction industry. BIM process is more than modelling buildings, and it provides the opportunity to drive efficiency and effectiveness to the information management of build projects. Accordingly, Building Information Models (BIMs), typically known as semantic three-dimensional parametric models, are fast becoming the comprehensive information source in Architecture, Engineering and Construction (AEC), and Facility Management (FM). The use of BIM in existing buildings has been hampered by the challenges and limitations surrounding the available technologies. The most popular and commonly used approach for generating models is to manually generate 3D artefacts utilizing point measurements extracted from range-based technologies (typically 3D laser scanning). In the recent past, several studies have been carried out to make the retro t BIM development process as effective and efficient as possible by developing different methods for mapping 3D models using Point Cloud Data (PCD) as the main source of information. However, an appropriate fully generated parametric model is still some way away. In this paper, we review the-state-of-the-art to address the research gap and challenges involved in generating parametric models before outlining the proposal of our approach. In this research, we employ Semantic Web technologies to capture parametric models. Elements are first recognized in PCD, and corresponding geometric information extracted from PCD are then tagged with Universally Unique Identifiers (UUIDs). Tags are then linked with the generated Resource Description Framework (RDF) data for each element. The core and challenging part of this research is the standardization process where RDF as a serialization is translated to Industry Foundation Classes (IFC) as a data model. The generated IFC format is then utilized to capture corresponding models. The primary results are very promising and should be of interest to the modelling of all kinds of building components, particularly historical building information modelling (HBIM).