To the average computer user, artificial intelligence seems to have emerged from nowhere – an amorphous and otherworldly force filling their social media feeds with dream-like landscapes and photo-realistic images of people who don’t actually exist.
For the people who spend their days designing and engineering buildings, however, AI and its close cousin, building information modeling, are technologies that have gradually been informing the entire development process.
Architecture tools
At Minneapolis-based architecture firm HGA, which has a Milwaukee office and has designed several Milwaukee buildings, AI tools are utilized to improve, and sometimes speed up, certain parts of the early building and site design process, said Jonathan Bartling, vice president of digital design at HGA.
AI programs or plugins like DALL-E and DALL-E-2, or Veras, can be “plugged into” traditional computer-aided design programs like SketchUp, Autodesk Revit or Rhinoceros 3D, to enhance or modify designs that have been manually rendered on a computer.
Veras, for instance, can alter the mood of a building rendering by showing it in different lights or seasons. Previously, that work would’ve had to have been done by a person using a more traditional design software.
Designers also sometimes use Autodesk Forma (formerly Spacemaker) to come up with options for building sizes and possible site layouts.
“It allows you to program the different building elements. You can tell it, ‘I want three buildings of these sizes for the site scale,’” Bartling said. “You are doing a lot more curating than iterating (with these tools).”
Building tools
At Milwaukee-based construction firm CG Schmidt, staff are mostly using predictive analytic AI tools, like Microsoft Power BI, that help their staff analyze data and make better decisions when it comes to choosing the best subcontractor for a certain job, or the best estimate for a specific project, said Brian Medina, the firm’s director of mechanical, electrical and plumbing and visual design and construction.
“It allows us to take a lot of data and sort through it fast,” Medina said.
The firm is also testing AI programs that other firms are using to help map mechanical layouts in buildings.
“These programs can help find the most efficient above-ceiling layout, like the insulation, plumbing and HVAC, within code required parameters and maintenance parameters,” Medina said.
At Madison-based construction firm Findorff, AI tools are mostly being used for budgeting and quality control purposes, but the firm is also aware of developing potential for design enhancements like those Medina mentioned. One way the firm is using AI is by employing an optical character recognition software, or OCR, to optically scan building plans to help analyze structural elements.
“At the beginning of every project, you are coming up with a budget, so you need to know how many windows, walls, how many doors are in those plans,” said Donnie Bogle-Boesiger, virtual design and construction lead at Findorff. “The software uses OCR to read those plans for you to help you with budgeting.”
The firm also uses AI in post-construction quality checks, using laser scanning of completed projects to see how accurately things were installed.
Embracing change judiciously
As AI design tools continue to grow in popularity and scope, the key to using them effectively, Bartling said, will be putting operational and philosophical guardrails in place to ensure the next generation of architects knows there’s a place for organic contributions.
“We have to make sure that we are not just replacing interns with AI. If there are tasks that are being replaced, we have to think about other ways of evolving talent,” he said. “There is a creative agency to the process that we don’t want to lose, by doing things, better, faster and smarter.”
For Medina, AI is coming along at a time when the construction industry is understaffed and in need of ways to streamline the process and recruit young talent.
“Construction is going to have to adapt. Younger workers are going to want more immediate results, because that is what they are used to,” Medina said. “We are not giving our younger generations credit. They don’t need to spend the first five years (of their careers) drafting.”