Computational and Generative Design
In recent years, the AEC industry has begun to identify a distinct methodology of computational techniques as “generative design”, an approach in which various types of inputs are computationally processed into one or more results of varying degrees of contextual suitability. Commercially, capabilities have at times outstripped professional discourse, so to support discussions on the topic I wrote a short LinkedIn article entitled “What is Generative Design?” where I defined the field as “…the automated algorithmic combination of goals and constraints to reveal solutions”. The coming changes in the AEC industry are rooted in the idea that automation can now enhance the traditional process of making decisions about buildings. Where it was possible to explore a few options for a design, engineering, or construction direction with the help of a few interns and hours of work, it’s now possible to explore hundreds or even thousands of possible solutions generated by scalable computation in a matter of seconds. Increasingly, building professionals, land developers, and owners recognize that far beyond merely recording decisions arrived at by other means, computing environments have the potential to both expand decision possibilities and help project teams arrive at better solutions.
Academic work in this area began more than 50 years ago, as captured in Gordon Pask’s 1969 paper “The Architectural Relevance of Cybernetics”, and the industry has been trying to realize that vision ever since. Advances in the last decade in both cloud technologies and computer programming paradigms finally place into the hands of building professionals the means to capture, scale, and deliver their expertise worldwide to the right people at the right time, for improved outcomes. Distributing expertise through computation has the potential to not only accelerate decisions but also to increase their accuracy and value to every building project.
Leaders in the AEC industry are beginning to adopt generative design methodologies and apply them to professional practice. To accelerate this trend, I’m very pleased to announce that DBEI’s BILT 2019 conference will be adding a Generative Design stream of classes to the North American event, which I’ve agreed to lead. We’ll be looking for abstracts tagged with “Computational Design” that cover research in this area but also that showcase the applied science of generative design in real projects. If you’ve developed automated methods for capturing and delivering your organization’s expertise, have applied machine learning productively to your firm’s data to arrive at better outcomes, or have found a way to combine goals and constraints algorithmically to reveal building solutions, we want to hear from you.
Submit your “Computational Design” abstracts here for review and let’s help the AEC industry advance to the next level of project delivery!