February 12, 2020

There has long been a gap between academia and design practice. Over the last few months, I’ve had the exciting opportunity to help lead a new partnership with the local academic community in Charlotte, N.C.

Merging Architecture and Data Science

During the fall 2019 semester, we created a collaborative sponsored studio course with Gresham Smith and the design computation program within the UNC-Charlotte College of Arts + Architecture. In this initial round, we engaged five students studying for their master’s degrees. Over 15 weeks, the students built prototypes, began testing their ideas and made a final presentation at our offices in Charlotte. Our overarching goal is to connect with the local academic community to help us learn how to develop tools and processes that can increase our capability to obtain, analyze and use data in our processes (design or otherwise).

Our hope is that this new partnership allows us to bridge the gap between academia and the realities of professional practice. Over the years, the tools architects use have become exponentially more versatile, quicker and powerful, but they’ve developed in ways that are no longer part of a typical architecture degree or practice. This partnership is a merging of architecture and data science; we’re bringing machine learning and algorithms into the design process with the goal of improving our processes, services and work product.

What We Learned: Leverage the Machine

What did we learn? Data science and its applications are an ever-expanding field, which will only increase in depth and presence in the coming years. Exploring ways to leverage data science processes and integrating them with design-thinking is one of the avenues that can help us anticipate future trends and needs and adapt successfully to what the future may bring. In addition, some of the students will spend their final semester furthering their proposed ideas in their theses.

I’d like to highlight one example that illustrates where we’re headed.

A student based their research on the impact machine learning can have on architectural design. Essentially, you take the architect out of the picture when it comes to design. Instead, you use a generative adversarial network (GAN), a class of machine learning systems, to facilitate the design process. Take the design of patient rooms in a hospital, for example. You can feed lots of images of patient rooms to a machine, and it can quickly generate thousands of possible iterations. In this example, the architect becomes a curator rather than a content generator.

As the student develops their thesis, we’ll partner with them by giving them access to our resources and our expertise, while they teach us their process and transfer the knowledge.

This is just the beginning. We’re using this partnership as a design lab that we can eventually bring to our practice. It’s just one of the ways we’re staying at the forefront of innovation in our industry.