This computational urban synthesis project develops a self-organizing spatial system for urban design. This design approach disengages urban design from the top-down notion of “master plan.” Instead, it proposes that the interaction of local decisions across multiple parallel conditions (spatial, financial, environmental, social) produces a complex system much closer to real-world cities. The imbalance and diversity characterizing local decisions suggests not a singular solution to the problem of urban design but a design system that accommodates multiplicity.

Algorithm Design

The design of the self-organizing system uses Cellular Automata (CA) as the primary computational method. Basic formulations of CA appear in the work of Ulam and von Neumann as well as of Conway and Wolfram. This project develops a novel formulation of CA to design a system that can more closely represent real-world cities.

In particular, the basic algorithm for CA distinguishes cells into two states: “solid” (black) and “void” (white). Then the rules determine how a cell changes its state based on its immediate neighbors. However, in this project the state of a cell transforms into multiple colors to represent alternative urban functions. In particular, each cell may be in one out of five states, each state corresponding to a distinct function. Examples of these functions are green areas, commercial and residential areas, and transportation and walking networks. The new 5-state CA algorithm also consists of a new set of rules that mimic constraints for urban design. Examples of these rules are the percentage of void cells that represent green areas, distances of residential area cells from transportation networks.

The algorithm functions in three-dimensions in an imaginary plot with dimensions 220mΧ330m and a spatial division into 5Χ5 m cells. This project uses 3DsMax and MaxScript for the implementation of the 5-state cellular automaton.

University: School of Architecture, Aristotle University of Thessaloniki
Course: 2S.302 Urban Synthesis
Year: Spring semester 2010-2011