Probobli Boboli | Florence, Italy | Unit 21 | 2023

This project explores the potential of using probabilistic simulations as part of a creative process for design, rather than simply a tool for validation, to further an understanding of the implications for designing with uncertainty at each step of the process. The exploration emerged from a historical study of the banks, libraries and fountains of Florence, looking at them as nodes in networked systems connected to reservoirs (of varying kinds) for managing uncertainty of different resources (financial, knowledge etc.) across time. The study goes back to the Medici rulers of sixteenth century Florence, who generated their wealth through banking, converted that wealth into power and influence over culture and material flows such as the water supply in the city and embodied it in the creation of the marble Neptune fountain in the Piazza Signoria. The interaction between water and stone as a slow process of weathering and erosion is a way into considering the design of a building across time considering the material flows through its life cycle.

The design process incorporates probabilistic models to distribute materials and channels across the site. It utilizes generative adversarial networks to simulate various stones for design synthesis. The resulting architectural elements disrupt and adhere to probable lines and channels, creating a unique blend of controlled design and random disturbances.

The project highlights the site’s invariants, such as trees and topography, and integrates pathways that connect buildings and focal points. It also includes subterranean spaces, a grotto, an olive grove labyrinth, a residence, and a library. These spaces incorporate flowing stone walls, meandering pathways, and computational tools to create dynamic and contemplative environments.

By employing probabilistic design principles and computational tools, this project reimagines the Renaissance gardens and explores the interplay between man-made design and the chaotic forces of nature, and what this suggests about the relationship between architecture, probability, and computation.