Visual Structuring for Generative Design Search Spaces

Abstract

With generative design strategies, the act of design, the problem-solving process, and the interaction, understanding and representation of design artefacts have changed. As such strategies entail a large number of possible design solutions, they also expand the design search space immensely. In addition, due to the automated design generation process, the interaction of the designer through the design process has decreased. This research problematizes two major points as (1) the broad design spaces of generative design systems that are populated by many design instances and (2) the elimination of designerly decisions and evaluation from the design artefact due to automated generation processes. As a solution to these problems, a visual structuring strategy is proposed. This strategy aims to act as a mediator between the designer and the design artefact during the generative process and manage the complexity resulting from the multiplicity of solutions. For visual structuring, three methods are proposed and classified as (1) perception-based visual structuring, (2) retrieval-based structuring, and (3) optimality-based structuring within the scope of this research.