A data-driven multi-objective approach to improving the quality of the urban space: The character of Sydney's public domain

This article investigates the quality of urban spaces, emphasizing the need for systematic quantification to guide design and policy decisions towards sustainable urban development. It highlights significance of the high-quality urban environments and the opportunities for translating quality as a measurable metric in the design stage. The study builds on previous research that identifies the key variables and percentages to calculate the four qualitative traits of beauty, ambience, safety, and comfort; identified through the analysis of urban suburbs in Sydney, Australia. The current phase of the research advances this work by using the calculated qualitative metrics as design objectives in the development of urban spaces through a generative multi-objective evolutionary algorithm (MOEA). This approach aims to integrate quantitative assessments into urban design, offering insights into creating inclusive and high-quality urban areas. Moreover, the research also proposes a method for integrating AI as a visualisation tool to convey the design output to the end user. The research underscores the complexity of urban quality, incorporating physical, social, and psychological dimensions, and advocates for a numerically driven approach to urban planning and design practices. • The quality of public space is embedded in the design process. • Safety, Ambience, Beauty and Comfort are design goals that are quantified. • The quality of public space is a measurable metric that improves the urban domain. • Ai is a suitable tool for design community engagement.