Architectural heat maps: A workflow for synthesizing data

Abstract

Over the last 5 years, large-scale 'data dumps' of architectural production have been made available online through project-specific websites (mainly compettions) and architectural aggregation/dissemination sites like Architzer, Suckerpunch, and Archinect. This reinforces the broader context of Ubiquitous Simultaneity, in which large amounts of data are contnuously updated and easily accessed through a dizzying array of mobile devices. This condition is being exploited by sports leagues and financial speculators through the development of tools that collect, visualize, and analyze historical data for the purpose of producing speculative predictive simulations that could lead to strategies for enhanced performance. We explore the development of a workflow for deploying computer vision, SIFT algorithms, image aggregation, and heteromorphic deformation as a design strategy. These techniques have all been developed separately for various applications and here we combine them in such a way as to allow for the embedding of the historical and speculative artfacts of architectural production into newly formed three-dimensional architectural bodies. This work builds on past research, which resulted in a more two-dimensional image-based mapping and translation process found in existing imaging protocols for projects like Google Earth, and transitions towards the production of data-rich formal assemblies. Outliers and concentrations of visual data are exploited as a means to encourage innovation within the production of architecture.