BIM-based graph data model for automatic generative design of modular buildings

As volumetric modular construction is gaining attention worldwide, there is a growing interest in the representation and model-based data exchange for modular buildings to optimise and communicate the design generation to prefabrication. This paper presents a Building Information Modelling (BIM)-based graph data model for the theoretic representation of essential characteristics in modular buildings, intending to automate and optimise the generative design for modular construction. The proposed new method involves the establishment of Industry Foundation Classes (IFC) model view definition (MVD) to comprehensively describe the required information for designing modular buildings. Following this, the IFC MVD is mapped to a graph data model, which includes the explicit formulation of absent spatial attributes, topological relationships, geometries, and semantics for volumetric modules. In compliance with the graph-theoretic representation, data transformation algorithms are developed to convert the required information in BIM to the graph model for assorted information query. Provided the extracted information, a prototype system is developed and examined via an illustrative example focusing on the automatic generative design of modular buildings. • Present a BIM-based graph data model for representing characteristics of modular buildings • IFC model view definition enhances the standardisation of precast design information • Explicitly formulate the spatial, topological, and geometric entities for volumetric modules • Develop new algorithms to automatically transform BIM to graph-structured data • Develop a prototype system to support the automatic generative design of modular buildings