Skip to main content

Data-driven design

Communicating a design is difficult. The use of code, scripts and parametric descriptions is a potential alleviation in this process. Codes are clear, precise, readable descriptive algorithms that are potentially accessible across the entire design team and can serve as future reference. In metaphysics natural phenomena are described by formulas. While this approach has proven to be difficult to adopt for the explanation of the greater universe, it is a valid approach in building design. Here, we harvest building information in formulas and explore various shapes as iterations thereof.

Computer-based design tools allow us to define forms via formulas and iterate their topology and geometry by minimal means. Compared to manual modelling techniques, this approach is not only highly accurate but can be extended to serve as a base for the further planning process. Information models are comparable to spreadsheets in their ability to capture the comprehensive data of a building. In such a process, the geometric model becomes a visual interpretation of this dataset.

Through our explorations of information modelling strategies, we aim to reduce complexity to an absolute minimum while maintaining the highest possible level of control over the output geometry. Simple mathematical equations such as trigonometric functions can describe a vast variety of forms. In building design, their simplicity reduces the complexity of the geometric model and allows for a more direct link to mature planning stages such as area calculations, facade penalization and fabrication. For example, in the process of designing high-rise buildings, we can use mathematical information models to determine the dimensions of the surfaces and slabs of a high-rise structure.

Often parametric design software prohibits collaborative design exploration. With increasing complexity, interaction with the information model becomes increasingly difficult. By developing custom building blocks within a parametric design framework, it is possible to overcome these hurdles in early design stages. By defining clear interfaces and data handling strategies throughout the digital design process, we promote an alternative approach to the widely implemented “ building information model  by using custom codes and programs to aid early design developments.

Deutsch, Randy (2011): BIM and integrated design. Strategies for architectural practice. Hoboken, N.J: Wiley.

Meredith, Michael; Sakamoto, Tomoko (2008): From control to design. Parametric. Barcelona: Actar-D.

Woodbury, Robert Francis (2010): Elements of parametric design. London: Routledge.

Burry, Mark (2011): Scripting cultures. Architectural design and programming. Chichester, UK: Wiley.