Although computers have significantly impacted the way we design buildings, they have yet to meaningfully impact the way we evaluate buildings. In this paper we detail two case studies where computation and machine learning were used to analyze data produced by building inhabitants. We find that a building’s ‘data exhaust’ provides a rich source of information for longitudinally analyzing people’s architectural preferences. We argue that computation driven evaluation could supplement traditional post occupancy evaluations.
My coverage of SmartGeometry 2016 for Architect Magazine, which featured a lot of robots and talk of building in space.
The first look at how WeWork’s R&D group measures architectural success.