Automating Urban Design?

As part of a series of lectures at MIT titled “What does urban science have to say about urban design?” Mike Batty presented a quantitative approach to person-to-person interactions in the pursuit of consensus about spatial planning, research that he believes can be used to optimize the (re)design of cities. By breaking down the design process into three parts—first the group dynamics, second the site(s), and third, the process of reaching consensus—he creates a neural net-like structure that correlates the opinions of the actors/stakeholders (residents, developers, corporations, etc) with the sites/buildings/locations under consideration (banks, housing, historic buildings, etc) to produce a series of optimized maps. Batty stated that: “The process of transmission between the actors can be changed quite dramatically so for example you might find that in certain iterations some actors don’t compromise etc, so there might be a situation where the network changes … that was the implication of the neural nets - the network might change over time so the network”.

Urban planners, watch out.