BIM and Semantic Enrichment Methods and Applications: A Review of Recent Developments

While considered a relatively emergent area of research (Bloch and Sacks in Autom Constr 91:256–272, 2018), Semantic Enrichment (SE) and Semantic Web services are among the prominent topics and trends in BIM research (BIM handbook a guide to building information modeling for owners, designers, engineers, contractors, and facility managers, John Wiley & Sons Inc, New Jersey). SE computational approaches provide valuable means to overcome current BIM limitations such as interoperability, topology relationships, extensions to standard schemas, among many others. Therefore, the study herein consists of a semi-systematic literature review on BIM-based SE systems and applications developed during the last decade. The article describes the computational methods and approaches identified, a classification of the screened papers according to their primary BIM Use ( https://bim.psu.edu/uses/ ), as well as reported limitations and recommendations for future developments. From the selected articles, main developments in SE techniques encompass multidisciplinary approaches comprising the use of Semantic Web technologies; inference rules, and rule processing engines; artificial intelligence methods; ontology mapping and semantic similarity; application of Industry Foundation Classes (IFC) libraries; and custom plugins. Considering BIM Uses, research was mostly focused on "Capture Existing Conditions" and "Validate Code Compliance". Other identified BIM Uses " verified three or fewer occurrences. Reported limitations state that more user-friendly interfaces are required to handle SPARQL queries (Lee et al. in Autom Constr 68:102–113, 2016). Moreover, the ontologies development process is deemed as time-consuming (Zhong et al. in Build Environ 141:127–142, 2018), and constraints were identified when trying to share semantic data between BIM and Geographic Information System (GIS) platforms (Zhong et al. in Build Environ 141:127–142, 2018). Future research may be expected in transitions to a more extended BIM paradigm, such as the formalization of Digital Twinning processes; discussions concerning a standard format for SPARQL query results (Karan and Irizarry in Autom Constr 53:1–12, 2015); as well as enhanced support through the transition to Linked Data and ontology-based systems.