An exploration of computational approaches to support creativity in design

Camilo Cruz - PhD Completion Seminar

Hansen Yuncken
Room, Level 4, MSD Building

This research explores ways in which computation methods can be used to support creativity in the design of environments for human inhabitation.

Design is a knowledge-driven, exploratory activity, aimed at transforming an existing condition into a desired one. Designers seek satisfying solutions given the constraints imposed by the context, the availability of physical and cultural resources, and the overarching objectives. Computational methods in design have resulted in improved efficiency. However, the capacity of such methods to contribute to the creative process – understood as the ability a designer has to find novel solutions embedded in the `design search space’ – is less evident. Recent advancements in artificial intelligence research have succeeded in identifying effective, and often unforeseen solutions for a given problem, especially when navigating large, complex search spaces. Consequently, this study examines the efficacy of advanced computational models to support the creative process in design.

Firstly, this thesis describes the development of a model to represent the design process and its outcomes in a systematic manner. This model, inspired by the General Design Theory and the Function-Behaviour- Structure model, incorporates evolutionary dynamics encapsulated within the design process. This model provides the framework for the implementation of computation methods for design. Secondly, the design, implementation and analysis of computational design methods are examined. A rule-based dynamic model and variations of evolutionary algorithms (including, interactive and multi-objective models) are described. The advantages and shortcomings of the methods are discussed in detail. Finally, recommendations for the application of computational methods to support creativity in design, as well as a theoretical proposition for a systematic understanding of the outcomes of design, are listed.

Biography:

Camilo is an architect and urban designer from Santiago, Chile. Over the past ten years he has worked in architectural design, visualisation and education, in Chile and Canada, mainly focusing on the use of computational methods for design exploration. Camilo is currently a PhD candidate in the Faculty of Architecture, Building and Planning, at the University of Melbourne.

Chair:

Dr. Fjalar de Haan

Supervisor/s:

Professor Justyna Karakiewicz

A/Professor Michael Kirley

Advisory Committee:

Professor Mark Burry

Dr. Fjalar de Haan