I recently competed another competition. This one was for a multi-use building in Fargo, North Dakota. The results were not what I would call interesting, but I was very happy with my own results. I worked on this project mostly on my own, with some help from Dennis Park and Daniel Yep Taboada.
There are two main ideas that I tried to work on in this project. Connectivity and distribution. These came together under the general topic of reticular networks. These networks are connective tissues that form fine meshworks around bodily organs. For this building I was interested in how to combine and connect the different programs of office, housing, retail, and public spaces.
The programs became the organs inside the body of the building (not sure how that relates to Deleuze and his body without organs). The connective tissue became a diagrid structure that wrapped around different zones. At certain points these zones are joined by bridges.
The distribution of the program was done using cellular automata. This is mathematical process that has gained some attention within computational design. It is a rules based system that are governed by on/off states of neighboring cells. This can be used to control density and create interesting patterns.
Here I used two sets of rules. One for the base where the density needed to be higher and a different set of rules for the towers. Based on different starting configurations and variations on the rules I created a large repertoire of options from which to develop a final design.
The seeming random patterns created by the cellular automata were perfect fit for my desire to develop public spaces through out the site. Different opportunities presented themselves as the units varied on different levels. These spaces emerge from the process and provide the type of unexpected results that computational design can achieve.
Another aspect of cellular automata I wanted to take advantage of is its modularity. Because it is based on repeating of the same unit through multiple iterations, I wanted to combine this with modular apartment types. After developing several apartment types I wanted to see how these could fit into the the structure produced by the cellular automata towers.
This is one of the aspects that I would have liked to work on more. It would have been nice to run this through an optimization sequence, where the different towers are evaluated to see which meets parameters based on programmatic needs such as how many 2 bedroom or 3 bedroom apartments are in each. Because of the short time frame of working on a competition I was not able to achieve this. Another aspect that I would have liked to work on is to develop the diagrid envelop into a manifold surface that moves through the open public spaces. This would have not only created a greater degree of connectivity, but also blurred the relationship between inside and outside and public and private spaces. but overall I think this is strong project that opens many avenues for further research.