Show simple item record

dc.contributor.advisorEastman, Charles
dc.contributor.advisorGentry, Russell
dc.contributor.advisorGoel, Ashok
dc.contributor.advisorHaymaker, John
dc.contributor.advisorShaefer, Dirk
dc.contributor.authorBernal Verdejo, Marcelo
dc.date.accessioned2016-08-22T12:21:54Z
dc.date.available2016-08-22T12:21:54Z
dc.date.created2016-08
dc.date.issued2016-05-23
dc.date.submittedAugust 2016
dc.identifier.urihttp://hdl.handle.net/1853/55571
dc.description.abstractThe general problem that this research addresses is that despite the efforts of cognitive studies to describe and document the behavior of designers in action and the evolution of computer-aided design from conceptual design to fabrication, efforts to provide computational support for high-level actions that designers execute during the creation of their work have made minimal progress. In this regard this study seeks answers to the following questions: What is the nature of design expertise? How do we capture the knowledge that expert designers embed in their patterns of organization for creating a coherent arrangement of parts? And how do we use this knowledge to develop computational methods and techniques that capture and reuse such expertise to augment the capability of designers to explore alternatives? The challenge is that such an expertise is largely based on experience, assumptions, and heuristics, and requires a process of elucidation and interpretation before any implementation into computational environments. This research adopts the meta-modeling process from the model-based systems engineering field (MBSE), understood as the creation of models of attributes and relationships among objects of a domain. Meta-modeling can contribute to elucidating, structuring, capturing, representing, and creatively manipulating knowledge embedded in design patterns. The meta-modeling process relies on abstractions that allow the integration of myriad physical and abstract entities independent from the complexity of the geometric models; mapping mechanisms that facilitate the interfacing of a repository of parts, functions, and even other systems; and computer-interpretable and human-readable meta-models that enable the generation and the assessment of both configuration specifications and geometric representations. For validation purposes three case studies from the domain of customs façade systems have been deeply studied using techniques of verbal analysis, complemented with digital documentation, for distilling the design knowledge that have been captured into the meta-models for reutilization in the generation of design alternatives. The results of this research include a framework for capturing and reusing design expertise, parametric modeling guidelines for reutilization, methods for multiplicity of external geometric representations, and the augmentation of the design space of exploration. The framework is the result of generalizing verbal analyses of the three case studies that allow the identification of the mechanics behind the application of a pattern of organization over physical components. The guidelines for reutilization are the outcome of the iterative process of automatically generating well-formed parametric models out of existing parts. The capability of producing multiple geometric representations is the product of identifying ae generic operation for interpreting abstract configuration specifications. The amplification of the design space is derived from the flexibility of the process to specify and represent alternatives. In summary, the adoption of the meta-modeling process fosters the integration of abstract constructs developed in the design cognition field that facilitate the manipulation of knowledge embedded in the underlying patterns of design organization. Meta-modeling is a mental and computational process based on abstraction and generalization that enable reutilization.
dc.format.mimetypeapplication/pdf
dc.language.isoen_US
dc.publisherGeorgia Institute of Technology
dc.subjectMeta-modeling
dc.subjectKnowledge-based design
dc.subjectModel-based engineering
dc.subjectDesign cognition
dc.titleMeta-modeling design expertise
dc.typeDissertation
dc.description.degreePh.D.
dc.contributor.departmentArchitecture
thesis.degree.levelDoctoral
dc.date.updated2016-08-22T12:21:54Z


Files in this item

Thumbnail
Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record