Speculative urban types: a cellular automata evolutionary approach

The accelerated rate of urbanization in China is the motivator behind this paper. As a response to the observed monotonous housing developments in Suzhou Industrial Park (SIP) and elsewhere our method exploits Cellular Automata (CA) combined with fitness evaluation algorithms to explore speculatively the potential of building regulations for increased density and diversity through an automated design algorithm. The well-known Game of Life CA is extended from its original 2-dimensional functionality into the realm of three dimensions and enriched with the possibility of resizing the involved cells according to their function. Moreover our method integrates the "social condenser" as a means of diversifying functional distribution within the Cellular Automata as well as solar radiation as requested by the existing building regulation. The method achieves a densification of the development from 31% to 39% ratio of footprint to occupied volume whilst obeying the solar radiation rule and offering a more diverse functional occupation. This proof of concept demonstrates a solid approach to the automated design of housing developments at an urban scale with a ,yet limited, evaluation procedure including solar radiation which can be extended to other performance criteria in future work.