Integration of knowledge-based and generative systems for building characterization and prediction

  • Authors:
  • Ajla Aksamija;Kui Yue;Hyunjoo Kim;Francois Grobler;Ramesh Krishnamurti

  • Affiliations:
  • Tech lab, perkins+will, chicago, illinois, usa;School of architecture, carnegie mellon university, pittsburgh, pennsylvania, usa;Department of civil and environmental engineering, california state university, fullerton, california, usa;Us army corps of engineers construction engineering research laboratory, champaign, illinois, usa;School of architecture, carnegie mellon university, pittsburgh, pennsylvania, usa

  • Venue:
  • Artificial Intelligence for Engineering Design, Analysis and Manufacturing
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper discusses the integration of knowledge bases and shape grammars for the generation of building models, covering interaction, system, and implementation. Knowledge-based and generative systems are combined to construct a method for characterizing existing buildings, in particular, their interior layouts based on exterior features and certain other parameters such as location and real dimensions. The knowledge-based model contains information about spatial use, organization, elements, and contextual information, with the shape grammar principally containing style rules. Buildings are analyzed and layouts are generated through communication and interaction between these two systems. The benefit of using an interactive system is that the complementary properties of the two schemes are employed to strengthen the overall process. Ontologies capture knowledge relating to architectural design principles, building anatomy, structure, and systems. Shape grammar rules embody change through geometric manipulation and transformation. Existing buildings are analyzed using this approach, and three-dimensional models are automatically generated. Two particular building types, the vernacular rowhouse and high-rise apartment building, both from Baltimore, Maryland, are presented to illustrate the process and for comparing the utilized methodologies.