A semantic and rule based technique and inference engine for discovering real estate units in building information models

The construction sector is a project-based industry characterised by informative-intensive processes. Project information is composed by several documents among which graphical drawings traditionally represent a fundamental component. The fragmentation of this information along the construction process can be addressed with the use of Building Information Modelling (BIM). One of the fundamental components of BIM is the development of information models. These models contain a defined and shared semantic structure that allows the simulation of real entities in a digital environment. However, the quality of this simulation is bounded by the limits of the instruments used for its development. The work presented in this paper is the development a rules-based system that makes inference upon ontologies obtained from IFC models to recognise real estates units and their destination of uses starting from building information models. The proposed application aims to overcome the limits imposed by the IFC schema focusing both the reasoning process and the resulting enriching information on ontologies.