Swarm planning : development of generative spatial planning tool for resilient cities

Abstract

In dealing with unexpected impacts of climate change current spatial planning tools are irresponsive and inflexible. The outcomes of applications of these tools are very limited in number, producing static plans that if implemented are very vulnerable to climate hazards. Therefore, an innovative generative tool has been developed to support spatial planning which results in designs that are responsive and adjustable to unexpected, simulated changes. The development of the generative tool is informed by swarm planning theory, and by contemporary generative approaches in urban design and planning. The generative tool is modeled as an Agent-Based System and utilizes versions of the canonical flocking algorithm. The agents are abstract cubical units of space that represent building envelopes. The agents exist and work within an environment that represents a site in terms of topography, land value, and available/buildable land. The agents receive information from the environment and act upon this information. The unexpected climate impact is a simulated flood, which affects both the environment and the agents. The outputs of the tool are generated 'bottom-up' in order to study emergent spatial configurations, as massings of building units.