Designing Computational Systems

8/14/2013

E/Ye Design

One thing I’ve never really understood about computational design is that in most cases, it is used for a one-off. The computational system is created for one large project and quickly forgotten after the project is complete. While the current flavor of computational design is certainly in its infancy, the key strength of computation is its ability to be reused. This idea is highlighted by E/Ye Design’s development of a construction technique to create manipulated planes, and its use in multiple projects. In these cases the computational logic behind each form may be slightly different, but they are ultimately informed by the same basic problem solving method.

The reuse of techniques, however, is different in E/Ye’s large projects. As explained, the cost of creating these forms is not that different from any other large project because everything is custom. It’s the opposite of the approach that they utilized in their smaller projects, in which the construction technique partially determined the form. In these projects the overall design was determined computationally, and the repetition occurs in the details, that while custom, are all dealt with in the same way.

These two approaches to computational design don’t represent two opposing views of how to create architecture, but rather, they point to the different ways that architects think about systems design. In the bottom-up approach, the construction technique is developed first, and each design is a variation of one system. The joints and relationships between building components are constrained and as a result the system can be varied to suit different sites. In the bottom-up approach, the same construction logic is applied and adapted to different design problems. In this approach, the detail gives meaning to the form.

Conversely, in the top-down approach, the overall form is determined by calculation, and it is the details that are variations of one another. Each building is not necessarily a variation of any previous detail or design logic because the building is unique enough that each requires its own system. It represents a situation in which each problem (building) has vastly different design requirements guiding its design, whether it is solar, wind, or structural strength. Given a wildly different set of design goals, each computational system must be tailor made to create a successful form.

Ultimately the decision to use a bottom-up or top-down approach comes down to the economies of scale. When a bottom-up approach is used, the repeated use of the same construction technique makes each successive use cheaper, whereas from a top-down approach the scale of the project keeps the cost of each detail in the design low allowing for a custom solution.