Data-driven Urban Design

Abstract

Nicholas Negroponte and MIT’s Architecture Machine Group speculated in the 1970s about computational processes that were open to participation, incorporating end-user preferences and democratizing urban design. Today’s ‘smart city’ technologies, using the monitoring of people’s movement and activity patterns to offer more effective and responsive services, might seem like contemporary interpretations of Negroponte’s vision, yet many of the collectors of user information are disconnected from urban policy making. This article presents a series of theoretical and procedural experiments conducted through academic research and teaching, developing user-driven generative design processes in the spirit of ‘The Architecture Machine’. It explores how new computational tools for site analysis and monitoring can enable data-driven urban place studies, and how these can be connected to generative strategies for public spaces and environments at various scales. By breaking down these processes into separate components of gathering, analysing, translating and implementing data, and conceptualizing them in relation to urban theory, it is shown how data-driven urban design processes can be conceived as an open-ended toolkit to achieve various types of user-driven outcomes. It is argued that architects and urban designers are uniquely situated to reflect on the benefits and value systems that control data-driven processes, and should deploy these to deliver more resilient, liveable and participatory urban spaces.