Full Article:

The paper addresses the concept framework, design and fabrication process of an artefact exhibited at the Venice Biennale 2023. The artefact results from a workshop with architecture, civil engineering, and design undergrad students. The concept spun around exaptation as a possible strategy for resilient landscapes – in our case, in the context of favelas. The object expresses a machine reading of favelas to create an abstract landscape from its materials. The landscape was obtained with a pipeline of unsupervised learning algorithms and parametric design that read images from favelas, extracted their features, and translated their latent space into geometry. The paper focuses on the methodology without neglecting a critical view of conceptions and meanings of favelas. It covers four parts: 1. Discuss the preeminent concepts; 2. Detail the design phase steps and methods; 3. Describe the fabrication techniques; 4. Discuss the results and exaptation from the favela’s point of view.