Brave New World: AI gains greater role in construction industry

From reality capture to modeling site development options to searching data lakes for relevant project comparisons, artificial intelligence (AI) is playing bigger and more varied roles inside the construction industry.

During BC&E’s Understanding AI in Construction webinar in August, representatives of construction, design and technology companies discussed the tasks that AI is already performing in the industry, possible impacts of AI on the workforce and best practices for implementing AI in companies. 

“AI is no longer a futuristic concept,” said Daniel Audette, Associate, Technology in Practice Specialist with GWWO Architects and moderator for the webinar. “It is here, it is growing rapidly and it is impacting the way we design, build and manage construction projects.”

Those impacts can be seen in multiple functions.

The reality capture technologies at OpenSpace enable a construction worker to put on a hardhat outfitted with a 360 camera, walk a construction site and capture two, 360-degree images per second, said Adam Settino, Senior Manager and Solutions Engineer at OpenSpace. Within 20 minutes, the technology will plot all images to a construction plan, enabling crews to assess as-builts and progress over time.

When he started working on stormwater management and other infrastructure projects, Brett Settles said it would take him two days to generate five options for laying out a site. Using SITEOPS, workers can now complete that task in two hours and generate two million options, said Settles, Product Owner at Allied BIM.

Employees at Harkins Builders today can utilize a large language model (a type of AI) to help compile project bids. They can ask the system to search out all similar projects the company has completed for similar clients and architects, compile data on each project’s financials, and flag the risks presented by such projects, said Patrick Hennessy, Director of Scheduling and Analytics at Harkins.

Successful implementation of AI-enabled technologies, however, is not simply a plug-and-play activity. 

Harkins Builders spent about six years optimizing its data management systems within a central, data lake before it was prepared to utilize a large language model. The company is now working on leveraging AI to improve project controls and procedures, especially scheduling. 

While technology firms are steadily generating new AI applications for construction, those technologies are still evolving and many are developed by people who do not have construction-industry experience.

Consequently, construction and technology professionals alike say collaborations between those two industries are essential to successful implementation of AI in construction.

“One of the best ways we can understand AI and figure out how it works with our processes is by partnering with product developers,” said Carlos Zuluaga, Virtual Design and Construction Manager, Harkins Builders. “Platforms like OpenSpace, Drone Deploy and others are trying to develop tools. When we provide feedback of actual cases in a jobsite, it really helps them develop a better product.”

“Don’t be afraid to reach out to the software providers that you work with, ask them what they are doing on the AI product side and then ask if you can get involved in Beta testing, in providing feedback,” Hennessy said. “They will absolutely love you and you might even get a discount on your next renewal.”

Providing feedback helps developers better understand builders’ pain points. Sharing data, such as scheduling files on a completed project, helps developers better “train their [AI] models,” Hennessy said. 

Successful adaptation of AI-enabled technologies, however, also means addressing changes in the construction workforce.

“There is this dream out there that AI can do everything. No, it can’t,” Zuluaga said. 

While AI can quickly assemble and partly analyze data, humans still need to complete a substantial amount of analysis and decision making, he said. But together, the technology and human effort can make company operations more efficient. 

“On the design side, my mantra has been AI will not take your job but someone who knows how to use AI will,” Audette said. 

In a recent interview, the CEO of LinkedIn “mentioned that generative AI is likely to change most jobs skills by about 60 percent in the next five years,” Hennessy said. “He didn’t say it is going to take 60 percent of jobs away but it is going to change the job skills.”

Consequently, employers will need to help employees adjust to new technology adoption and the automation of some tasks. They will also need to develop plans to divert the time and resources saved through the adoption of AI into other productive tasks. 

AI not only presents the prospect of boosting the efficiency and productivity of construction companies, Zuluaga said. It also creates opportunities to expand the construction workforce by recruiting different types of employees. Rather than focusing solely on recruiting engineers, project managers and tradespeople, “maybe we need to hire a data scientist or somebody who has mechanical fabrication knowledge to help with our MEP installations.”

Bringing virtual reality devices, BIM models and other construction technologies to university job fairs is already beginning to attract computer science students to the construction industry, Settles said. 

“These are people who didn’t even know that construction had this sort of capability or that they even had a place in this industry,” Settles said. “AI is going to allow companies to keep capacity while they figure out how to bring new people into the industry.”