Generative Urban Design Concepts and Methods: A Research Review

Abstract

Generative design is a fast-growing design approach that is both important and pervasive. This approach has developed after the great digital revolution in the last few decades, especially during the first two decades of the third millennium. Taking advantage of the computational capabilities of modern computers, generative design offers designers the ability to test multiple and complex alternatives by automating design processes using algorithms. The automation makes it possible to manage an ever increasing number of variables and parameters that are difficult to manage manually. This paper presents an attempt to classify the approaches of the generative urban design methodologies applied in the field today, with the aim of highlighting the gaps in the current practices. By reviewing the previous available studies that dealt with the idea of generative urban design, we find that they can be classified according to the approach they adopt in four categories: the first is the approach of deforming or distorting a basic shape proposed by the designer, the second is the approach stemming from the simulation of real urban growth, while the third is the approach of dealing with the urban form on the scale of building blocks without being concerned with the spatial network, and finally the fourth is the approach that emerges from analysing a realistic spatial structure and then proposing modifications based on the results of the analysis. From this review, the paper proposes a new approach resulting from the analysis of a realistic spatial structure to build a generic model that can be the starting point of a synthetic process to be integrated into a generative design methodology.