Through the undertaken research, Zurichness can be understood as a combination of facades drawn from a matrix of windows, and a varying ground floor of arcades, glass showcases, double heights, and so on. The decision to bring into focus Zurich’s angles leads to an understanding of the volume of the building and this shows how the façades connect with each other on different streets. Through the use of machine learning and artificial intelligence, the essence could be distilled as a result of prolongated sessions of GAN training, going through an abundant quantity of images of the city.
With this, it was possible to identify typical Zurich elements (window matrix, shutters, overhanging roof, arcades, glass showcases…), and depending on selected GAN images it was possible to either amplify them or reduce them. In the context of a post-pandemic world, the project focuses on the composition of the facade, used to define the volumes inside, in their proportions and programs. Groups of windows define the volumes, keeping the most interesting space that is the rounded angle as an XTRA space to be shared between two apartments. This XTRA space introduces 3 systems of circulation: the access to the apartments, the access to the additional space for the public, and then the connection of the private living spaces.