Rule-based compliance checking and generative design for building interiors using BIM

Abstract

Building Information Modeling (BIM) is becoming an integral part of the design, architecture, and construction process, as it can integrate all building data in an accessible digital representation that can be viewed in a 3D environment prior to construction. This supports the capability of evaluating a model against building codes or design rules to ensure that the building meets the relevant functional and safety requirements for occupants. However, building regulations are typically represented in natural language and, to date, they have not been created with regard to the digital BIM design process. To automate the design evaluation of a building model, this paper describes a simple, yet extendable, domain-specific language for computationally representing building interior design rules and a method for evaluating rules in this language against a BIM model.Furthermore, previous language-based model-checking approaches have not explored the applicability of their proposed languages beyond model-design evaluation. This paper demonstrates the use of our language in a second, more complex, use case, namely the automated generation of multiple valid alternative model interior designs, compliant to a set of design rules expressed in our language. We have evaluated the usefulness of our language in two generative-design case studies, namely kitchen layouts using a set of real-world kitchen design rules provided by an commercial partner, and living-room layouts using rules put forward in earlier automated-design research. Therefore, this research bridges the concepts of automated model checking and generative design relying on the same rule language.