Computed Density

Inner city housing using computational design and robotic fabrication

More Info
expand_more

Abstract

The current demand for living in cities necessitates architectural designers to reflect on how this densification is to be done. Based on analysis of a current densification project in Rotterdam (Rotta Nova) different problems with current densification designs are formulated. From these problems an alternative design for inner city dwellings is suggested, using the advantages of computational design and robotic manufacturing. The first problem with current densification strategies has to do with the bounding volume of the building, which does not take into account the negative impact on its surroundings (mainly with regards to shading). Secondly, the configuration of the apartments themselves does not take into account the limited space available. Ideally, in a high-density environment we would assign as much volume to every space that it actually needs; living rooms would desirably be spacious, while bedrooms can be smaller. Thirdly, current densification designs often replace green areas within the city. This greatly worsens the urban heat island effect, especially in Rotterdam. Based on these three problems I propose an alternative densification strategy, materialized as an alternative design for the apartment building I analyzed. This densification strategy consists of three scale levels. On the macro scale level, a bounding volume is generated which ensures minimal shading impact on surrounding buildings and public spaces. On the meso scale level a generative algorithm is developed which populates this bounding volume with apartment spaces of different sizes, with exactly the dimensions they require; functional areas are minimized, while living quarters are maximized. On the micro scale level, a building system is developed using concrete 3D printing which allows the production of the resulting non-standard building. These efforts result in an alternative design on the site of Rotta Nova. On one hand, this is site specific; the bounding volume of the generative model is based on local site conditions. However, the workflows on all three scale levels are applicable in other dense urban areas as well, be it with slight modifications.