AI-based tools can help us reduce mundane tasks and improve sustainability in building design
Most of us are already accustomed to using AI-based tools in our personal lives—whether that’s a movie recommendation on a streaming media service or directions on the map app on our phone. But what about in our professional lives? In architecture firms around the world, architects are still manually drawing walls, doors, and windows thousands of times, each time with subtle differences. Caoimhe Loftus, regional lead for CallisonRTKL digital group, shares how we can reduce the need for people to handle repetitive tasks using AI. She highlights ArchiGAN, a machine learning tool trained using a library of over 800 apartment types that can lay out rooms and position doors, windows, and furniture; and Architext, a tool that can generate apartment layouts based on words. But ultimately, the greatest value of AI in architecture may come from helping us make buildings more sustainable. She cites Spacemaker as an important tool that can analyze building designs for sun, noise, wind, and other factors while they are still in the conceptual phase, rather than wait weeks or even months for manual analysis. The role of the designer changes with these tools, she points out, becoming one of choosing the right prompts and curating the output, but it has the potential to make the job more fulfilling and creative—and the final outcome better for people and the planet.
About the speaker
Caoimhe Loftus is a regional lead at CallisonRTKL’s Digital Group, where she combines her experience as an architect and her passion for innovation to demonstrate how technology can improve the way we work. Her areas of expertise include automation, optimization, and new software adoption. Loftus holds degrees from Technological University Dublin and the University of Westminster.
Coimhe Loftus and Arne Bassoe-Eriksen demonstrate how to leverage Spacemaker to kick off a project and validate early design decisions, how to migrate the model into Revit for further design development, then how to go back into Spacemaker for additional validation if necessary.
Reducing a building's carbon footprint decreases its running costs, improves the workforce's sense of purpose, and raises property value. Daniel Gameiro and Justin Taylor show an outcome-based design workflow from schematic design in Spacemaker to design development in Revit and Insight.
DaisyAI is the first timber design software that produces optimal, code-compliant timber designs in minutes. Kratos is an Autodesk Research project using AI methods to rapidly evaluate many structural designs in multiple materials including concrete. This case study showcases the collaboration between the two tools, using Kratos to calculate load-bearing walls in a timber structure which Daisy can use as input to produce detailed floorplans. The final result? A house built in the UK.