PROCESSING COMPLEXITY / 18.05.11
Programme:
14:00 Wassim Jabi (Welsh School of Architecture) Opening comments
14:30 Michael Batty (UCL, CASA)
15:00 Break
15:15 Christian Derix (Aedas)
15:45 Marco Vanucci (Adams Kara Taylor)
16:15 Break
16:30 Simon Lannon (Welsh School of Architecture)
17:00 David Hines (Populous)
17:30 Break
17:45 Discussion
-
Introduction: Agent-based models for urban simulation and speculation
The traditional method of conducting urban analysis has been to use drawings to outline site assets and features, represent invisible or hard to detect forces, and document ephemeral phenomena. These site analysis maps are then augmented with textual information, diagrams, photos, and models. While these types of representations can be very useful in understanding certain aspects of a site or context, they fall far short in allowing us to understand the micro-based forces that act on it and collectively influence its organisation. In addition, they usually fail to explain the multitude of local low-level interactions that take place within it: Does this subway station design create congestion at the ticket stalls? Does this urban configuration create wind corridors? Will this location of commercial stores attract enough foot traffic? These types of questions are difficult to answer without the help of spatial systems that model the underlying topology of a site and temporal multi-agent-based systems that allow a series of agents/bots/cells to operate independently and/or collaboratively and with purpose within a set of constraints.
The advent of agent-based modelling and generative design systems has enabled researchers to use a new set of tools for understanding and modelling complex phenomena such as people movement, human settlement and city development. Furthermore, designers are increasingly turning to these systems to help them understand the context of their projects, visualise flows and forces that are next to impossible to document using more traditional methods, and speculate on new emergent spatial organisations that they detect from these systems. The fundamental underlying principle behind these methods is that higher-level systems of organisation can be better understood as emergent patterns of a multitude of lower-level local interactions of agents/cells/atoms. The seminar will discuss the state-of-the-art in the major themes that have developed over the last few years including: cellular automata, fractals, agent-based models, emergence and complexity theory, and temporal and spatial/topological models. The participants in this seminar include world-class experts in the field of spatial simulation that will outline the major themes and report on their research in using new paradigms of urban, spatial, and people movement analysis that rely on computational methods.
(Image Credit: Mechanical Attractors by Richard Marxer)
-
-
-
Supported by the Welsh School of Architecture, Design Circle RSAW South and CEBE
