Machine View of Boston

A Machine View of Boston presents a matrix of 30,000 buildings in central Boston organized by form.

It was products with the RoweBot, named for architect Colin Rowe, a tool programmed to use vision-enabled bots to scan and synthesize billions of figure-ground shapes and building plans. These shapes are then systematically categorized and reorganized based on relative geometric properties such as relational similarity, clustered affinity, and overall categorical distribution of architectural form and typology. The RoweBot produces cartograms (map-like representations) that do not chart geographic space, but instead the space of similarity, affinity, and perception, revealing, through its gaps, a terra incognita of design yet to appear.

Date: 2019
Clients: Boston Society of Architects
Team: Andrew Witt, Tobias Nolte