Machines “R” Us

Machines “R” Us is a fictional stadium generated from a machine learning model. This project investigates the way stadiums intervene with existing landscapes. By training a machine learning model on nearly 3GB of aerial imagery from existing stadia across the world, viewers can see and understand the latent ideas of beauty embedded within the fabric of society. These latent images coalesce around 3 main clusters of stadium typologies (with the given names): “The Mini-City”, “The Performance Island”, and “The Urban Chameleon”.

Not only does this project serve as a tool for understanding cultural preservation, but the project also offers visual feedback to conduct a more advanced spatial analysis around anticipatory zoning, typological ideals, and the lasting effects of stadium masterplans.

Two of the overarching questions this methodology addresses include, but aren’t limited to:

• How might we redraw urban grids through innovative stadium master planning?

• How might we make a stronger case for stadium masterplans that aid neighborhood development?

Project Type :: Environmental Design

Year :: 2022