Behavioral formation: multi-agent algorithmic design strategies

Snooks, R 2014, Behavioral formation: multi-agent algorithmic design strategies, Doctor of Philosophy (PhD), Architecture and Design, RMIT University.

Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Snooks.pdf Thesis application/pdf 58.90MB
Title Behavioral formation: multi-agent algorithmic design strategies
Author(s) Snooks, R
Year 2014
Abstract Emergence and the behavior of complex systems are increasingly defining the contemporary understanding of processes as diverse as natural phenomena, social structures and the existence of consciousness. An interest in this self-organising behavior is explored in my architectural work through the development of a computational, generative design approach. The focus of this thesis is the interaction of emergent processes of formation and architectural design intention: a conceptualisation that has emerged through a process of reflecting on my practice of architecture.

This PhD reflects on the projects, processes, techniques and concepts that have come to shape and define my practice. The body of work examined here has been developed through Kokkugia (an experimental research collaboration with Robert Stuart-Smith) and my architecture practice, Studio Roland Snooks. Through this process of reflection, I have extracted, articulated and developed a series of strategies, ideas and sensibilities that have been redeployed in subsequent projects, and point to future directions in this work.

This body of work is positioned within an emerging group of architects who are engaging with complex systems, generative design strategies and algorithmic techniques. The original contribution this design research makes within this milieu lies in my focus and experimentation with a process of embedding architectural design intention within generative algorithms. Design intention is recast as behaviors – discrete, micro-scale architectural decisions, relationships or procedures – that are encoded within multi-agent algorithms. It is the local interaction of these agents that self-organises architectural design intention at the macro-level. An iterative design process has been established that negotiates between this emergent process, subjective evaluation and direct design decisions. The multi-agent algorithms developed in this design research draw on the logic of swarm intelligence (a branch of complexity theory), which describes the emergent, collective behavior that can be found in phenomena such as schools of fish, flocks of birds, social insects and slime mould.

The implications of working through these highly iterative, non-linear, computational design processes are manifest as a compression of tectonic hierarchies and a blurring of geometric types. This behavioral design process negotiates between various architectural design intentions through the geometry of architecture, creating a synthetic but differentiated assemblage, and uncoupling geometric elements from architectural roles. The architectural projects that we coax out of these processes exhibit emergent qualities – strange, intricate characteristics; complex order; and intensive capacities to affect.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Architecture and Design
Keyword(s) architecture
algorithmic design
multi-agent systems
swarm intelligence
Version Filter Type
Access Statistics: 594 Abstract Views, 1375 File Downloads  -  Detailed Statistics
Created: Tue, 28 Nov 2017, 08:57:35 EST by Keely Chapman
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us