Algorithmic approaches toward information-based urban form

Abstract

This study explores the potential of using information flow as a spatial generator in urban architecture, through an algorithmic approach that emphasises relational thinking and adaptability in shaping the built environment. Bridging computational design and ecological paradigms, it reframes architecture as a medium for information exchange in the second machine age. The proposed method employs algorithmic processes Cellular Automata and Swarm Intelligence to generate dynamic spatial systems that address the logistical complexities of urban contexts. Applied to real urban sites, the CA-SI workflow reveals aggregation fields, directional gradients, and distributed trajectories that converge into five recurrent network topologies: annular, co-linear, co-vertex, diffused plane, and intersectional plane. These swarm networks prioritise relationships and interactions over isolated forms and foster a self-organising and metabolistic nature of urban networks. The findings indicate that information-driven spatial tendencies interact with and reorganise existing material conditions, which result in a co-constituted physical–digital field understood as multispace. This interplay highlights both the generative potential and interpretive limits of computational abstraction. By integrating modular and scalable components that dynamically respond to spatial and informational demands, the study offers a morphogenetic approach to urban architecture and positions information as a driver of adaptive and interconnected urban ecosystems.