Rule Language Based Automated Compliance Checking for Interior Generative Design Using BIM

• Author / Creator

  Sydora, Christoph

• Building Information Modeling (BIM) has become an integral part of the design process, as all building data is accessible in a digital representation and can be viewed in a 3D environment prior to construction. This supports the capability of evaluating or checking a model against building codes or design rules, imposed from construction codes to cultural preferences to the owners’ styles and aesthetics, a procedure necessary to ensure a building meets the functional and safety requirements for occupants. However, as building regulations are typically represented in natural language, to date they have not been created with regard to the digital BIM design process. Therefore, checking a model against these design rules is still a time consuming and error-prone task involving knowledgeable individuals reading rule documents and manually assessing a building design. Furthermore, design rules are subject to interpretation rather than structured for machine interpretation allowing for rule assessments to differ among individuals.

  To automate the design evaluation of a building model, this thesis describes a simple, yet extendable, domain-specific language for computationally representing building rules. We describe how implicit information is extracted from the BIM model, as necessary for the rule evaluation on the building model. Previous approaches to model-checking generally require experienced coding knowledge and have been tailored to meet specific building regulations with minimal support for rule creation.

  The model evaluation also provides opportunity for automatically generating multiple valid alternative solutions, compliant to the design rules. Due to time constraints, designers typically only explore a few options; computers can efficiently create a potentially much larger number of alternatives, which users can then compare to select the design that best suits their individual preferences, depending on cost, environmental impact, and aesthetics. Therefore, unlike the previous methods of rule-based model checking and automated layout designs, this thesis bridges the concepts of automated model checking and design using a single unified rule language.

  The research is evaluated on two particular instances of the general generative design problem. The first is the task of generating 3D kitchen layouts, based on a BIM model of the kitchen space, a product catalog of 3D models of kitchen furnishings, and a set of design rules. The generative-design method starts with an empty kitchen and implements a heuristic search of the solution space by incrementally selecting and placing a required item and checking the degree to which the resulting model complies with the given kitchen design rules. We have demonstrated the effectiveness of our method by comparing the designs it produces against a set of real-world kitchen examples, obtained from architecture diagrams available online.

  The second task is to generate a living room, using rules interpreted from previous literature on layout design that were not written in a language-based evaluation method. Using the same method, we create living rooms that meet the design requirements from the previous literature.

• Subjects / Keywords

  • Model Checking
  • Building Information Modeling
  • Generative Design

• Graduation date

  Fall 2019

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-8cwz-5a81

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.