| dc.description.abstract | Territory, as an incipient design setting, is progressively displacing conventional notions of site within design research and practice, and, with this, the design professions are increasingly exploring their agency as instruments of territorial intervention, formation and reformation; a disciplinary shift witnessed in recent discourses such as Landscape Urbanism, Ecological Urbanism, and Ecological Design. With this renewed contextual perspective, complexity is acknowledged as a base condition, accompanied by an operative shift toward geographical contexts, techniques, and representations which foreground systems-oriented perspectives with process-driven approaches. Similarly, a pivotal shift in focus from the essence of objects to the management of dynamic spatial systems is increasingly taking root.
Yet, the specific methods, tools and techniques used to operate within this expanding field of practice are deserving of further exploration in their own right, and it is this point that serves as the primary motivation for this thesis. As such, the thesis proposes a methodological framework which operates at the intersection of territorial design research and computational thinking, proposing the use of methods, techniques and tools drawn from spatial data mining, machine learning, and computational modelling as mechanisms for dealing with complexity in territorial systems.
The driving motivation in the development of this framework is to eliminate the gap between contextual analysis and the development of a design response, by exploring ways in which the data which is used to characterize a design context can be carried directly through to inform a design process. The framework, offered as a black-box system, is examined by way of a specific implementation, using historical data from the 2011 Japan Earthquake and Tsunami as the basis for a design experiment.
After exploring each phase of the framework – Discovery, Modelling, Formation & Exploration – the challenges and limitations of appropriating extra-disciplinary devices, and the role of subjectivity in computational modelling are discussed. Lastly, looking forward, a recursive implementation of the proposed framework is proposed as an avenue for future research and development. | en |
