Generative urban design with cellular automata and agent based modelling

Abstract

This paper reports on initial findings of a bigger research project that set out to explore the potential of generative algorithms in landscape architecture, urban design and architecture. The paper focuses on how urban morphologies of unplanned settlements can be modelled as emergent phenomena using parallel computing. Theoretically the research stems from Hiller’s discourse about space syntax, summarised in Section 2. The paper outlines the concepts behind generative design and illustrates the principles of Cellular Automata and Agent Based Modelling using some simple examples. The models are described in detail and their potential usefulness in design education is demonstrated. The potential of using such models in design practice is also evaluated.