A Generative System for the Terrain Vague

This work, through the study of the Terrain Vague impacted by the Transcarioca BRT in Rio de Janeiro, developed a four step methodology in order to re-discuss a sample of residual spaces that were identified and classified, then presenting an algorithm capable of translating and optimizing these spaces through urban parameters and physical criteria.

From the studied session, I perceived stagnant areas calling for intervention. Despite this express bus lane contributes to the mobility of the Rio suburb, it also reorganizes the city's urban fabric, bringing new challenges to the commoners. I asked myself how can technology — predecessor of countless changes, such as the BRT — be capable of regression? Expecting an answer for this question, this work hoped to resolve the premise that technology rediscusses the urban landscape, but rather incorporating the technology itself to engender an architectural development methodology, which consisted of four stages. Initially, through the context analysis, 99 were the number of areas passable for intervention, where I considered 50 of them as residual spaces — areas that were destroyed but had not been reconstituted — a scale of Terrain Vague. These areas had mutual characteristics that I translated into urban variables to my algorithm. In the next step, based on a visual grammar introduced by the work of ZAERA-POLO, A (2003), I organised the previous data to classify the selected cases by categories. Subsequently, in the third step, through the Rhinoceros/Grasshopper visual programming, I created an algorithm to generate volumetric solutions, associating space properties with geometric operations, each one with its specific rule; a) public spaces to get transformative operations and to become urban squares; b) semi-public spaces to get additive operations and to become stores or restaurants; c) semi-private spaces to get subtractive operations and to become cinemas or theatres. Finally, I also came with 3 physical criteria that evaluate shade, open space hierarchy and urban density of each solution in order to select the most optimized one. The intention of my work was not to solve the Terrain Vague issues, but to present an algorithm that allows optimized occupancy in those areas. It is still early to say if the algorithm works properly, but it is already indicating unexpected results due to the randomness of the operational processes, contributing to the development of a new evolutionary system.