Geolocation in the Semantic BIM
Muhammad Fahad, Nicolas Bus · 2019
Building codes are the rules and guidelines that specify the minimum acceptable level of safety, accessibility, general welfare, etc. of building models. Code compliance checking of building models requires precise identification or estimation of the real-world geographic location of IFC objects. Geolocation based on Bounding Boxes is a very simple technique defined by minimum and maximum longitudes and latitudes. In this paper, we discuss that bounding box technique cannot help precise identification of IFC objects and does not fulfill all our requirements of code compliance checking. Therefore, we have to go beyond bounding boxes and analyze Well-known Text format for the precise identification of data objects in the building model. Well-Known Text is a text markup language for representing vector geometry objects on a map and offers a precise and compact machine and human-readable representation of geometric objects. We explored two geospatial tools (Stardog and GraphDB) for working with Well-known Text and illustrate various examples narrating their similarities and differences on various test cases. Both the tools implement several extensions to ease queries for processing geospatial data. These Geo-SPARQL extentions are designed to accommodate systems based on qualitative spatial reasoning and systems based on quantitative spatial computations.