Stouffs, R., Krishnamurti, R. and Cumming M., Mapping design information by manipulating representational structures. G-CAD: International Symposium on Generative CAD Systems, Accepted, 12-14 July 2004, Carnegie Mellon University, Pittsburgh, USA.

RUDI STOUFFS1, RAMESH KRISHNAMURTI2 AND MICHAEL CUMMING1 1Delft University of Technology, Faculty of Architecture, Delft, The Netherlands
2Carnegie Mellon University, School of Architecture, Pittsburgh, PA, USA

Abstract. Design problems require a multiplicity of viewpoints each distinguished by particular interests and emphases. Alternative viewpoints necessitate different representations of the same entity, albeit a building or building part, a shape or other complex attribute. We argue that the exploration of alternative design views can be supported by providing access to the representational structure and by allowing the structure to be manipulated through incremental changes. Hereto, we briefly describe the representational framework of sorts and present its support for comparing representational structures and mapping design information according to it. We illustrate the creation and manipulation of structures and their comparison. We consider the specification of design queries through the integration of data functions into representational structures. We conclude with a presentation of future work.
Computational design relies on effective information models, for the creation of design artifacts and for the querying of the characteristics of such artifacts. Mäntylä stated in 1988 that these (geometric) representations must adequately answer “arbitrary geometric questions algorithmically.” Without emphasis on the geometric aspects, this remains as important today. However, current computational design applications tend to focus on the tools and operations for the creation and manipulation of design artifacts. Techniques for querying receive less attention and are often constrained by the data representation system and methods. Nevertheless, querying a design is as much an intricate aspect of the design process as is creation and manipulation. Design is also a multi-disciplinary process, involving participants, knowledge and information from various domains. As such, design problems require a multiplicity of viewpoints each distinguished by particular interests and emphases. For instance, an architect is concerned with aesthetic and configurational aspects of a design, a structural engineer is engaged by the structural members and their relationships, and a building performance engineer is interested in the thermal, lighting, or acoustical performance of the eventual design. Each of these views⎯derived from an understanding of current problem solution techniques in these respective domains⎯requires a different representation of the same (abstract) entity. Even within the same task and by the same person, various representations may serve different purposes defined within the problem context and the selected approach. Especially in architectural design, the exploratory nature of the design process invites a variety of approaches and representations.
Design views facilitate a visual inspection of design data and information. Design queries support the analysis of existing design information in order to derive new information that is not explicitly available in the information structure. Design views can be understood to be discrete and domain-specific; design queries on the other hand seem to indicate small incremental steps transcending common, domain-specific views in search of information that does not naturally form part of the design view. At the same time, the result of a design query, possibly presented in the context of other design information, may be seen to define a design view and, similarly, design views may be expressed through design queries.
The distinction between discrete and incremental views can also be related to the development of integrated data models. Integrated data models span multiple disciplines and support different views. These allow for various representations in support of different disciplines or methodologies and enable information exchange between representations and collaboration across disciplines. Examples are, among others, the ISO STEP standard for the definition of product models (ISO 1994) and the Industry Foundation Classes (IFCs) of the International Alliance for Interoperability (IAI), an object-oriented data model for product information sharing (Bazjanac 1998). These efforts characterize an a priori and top-down approach: an attempt is made at establishing an agreement on the concepts and relationships which offer a complete and uniform description of the project data, independent of any project specifics (Stouffs and Krishnamurti 2001). These efforts also mainly target software developers who can ensure compatibility of their own representation corresponding to a particular design view with the integrated data model.



  1. Leave a Comment

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: