Aana: Experiments In Artificial Aesthetics In Architecture

Aspects of architectural aesthetics are tractable to computable solutions. Cloud computing, Web-based image sources, and neural networks present new opportunities for computational models of human experience of architecture. Review of four recent research projects illustrates four computational ways to approach aesthetics: physiological, sensory, active, and constructive. The examples also illustrate the use of several techniques, including neural networks, algorithmic reasoning, symbolic reasoning, parametric modeling, Web-derived datasets, and synthetic datasets. A tentative framework for organizing work in artificial aesthetics in architecture can help facilitate expanding attention to this field. Commercial applications and potential social impacts could help address rebuilding programs that ameliorate stress on the built environment caused by climate change, population migration, and population increases.