Mapping boundaries of generative systems for design synthesis

Abstract

Architects have been experimenting with generative systems for design without a clear reference or theory of what, why or how to deal with such systems. In this thesis I argue for three points. The first is that generative systems in architecture are implemented at a skin-deep level as they are only used to synthesize form within confined domains. The second is that such systems can be only implemented if a design formalism is defined. The third is that generative systems can be deeper integrated within a design process if they were coupled with performance-based evaluation methods. These arguments are discussed in four chapters: 1- Introduction: a panoramic view of generative systems in architecture and in. computing mapping their occurrences and implementations. 2- Generative Systems for Design: highlights on integrating generative systems in architecture design processes; and discussions on six generative systems including: Algorithmic, Parametrics, L-systems, Cellular Automata, Fractals and Shape Grammars. 3- Provisional taxonomy: A summery table of systems properties and a classification of generative systems properties as discussed in the previous chapter 4- Conclusion: comments and explanations on why such systems are simplicity implemented within design.