Smart Foundations: Preparing for AI and ML Integration in Construction | July 24, 2024
Continuing our post-event coverage of the 2024 Construction Tech Conference in Chicago, we look back at one of the more highly anticipated sessions focused on artificial intelligence and machine learning (AI/ML). Understanding AI/ML has the potential to rapidly revolutionize the AEC industry by using the vast amounts of data collected throughout a project (both new and historical data), we must be prepared to anticipate and welcome the various capabilities of this technology. To wrestle with this idea, the audience heard from a wide range of speakers including an investor, technology providers and, ultimately, a general contractor. Panelists included Alan Espinoza, solution architect, Graitec; Ivy Nguyen, principal, Autotech Ventures; Dan Williamson, director of AI, Ryan Companies; and Joel Hutchines, CPO, Slate Technologies.
Before an organization can effectively implement AI/ML technology or principles, a few criteria must be addressed and understood. See a few key takeaways from their discussion below!
Avoiding AI for the Sake of AI
As is true with implementing any new process or technology, problem identification must be the first step. Especially when considering something "new" such as AI/ML, architects, engineers and contractors need to have strategies in place to teach their employees what the technology really is and how it will impact their projects. During the discussion, the panelists emphasized that having clear strategies opens the dialogue internally about what information is needed and how data is used to generate insights that may change the work performed throughout an organization.
However, it is not always easy to navigate what AI is and why you might consider adopting it. For example, you might be AI-ready in certain areas or functional groups, but that does not mean your whole organization is mature enough. Deeply understanding your unique pain points and reasons for choosing one solution over another will allow you to ask the right questions and, perhaps most importantly, manage expectations and create the fastest time to value (i.e. generating ROI).
Who Owns the Data, Security Concerns
Once a problem is identified, a major hurdle identified by our panelists is getting over the fear that they’ll lose their competitive advantage by plugging sensitive information into AI solutions. Despite that notion, much of the "proprietary" knowledge is locked in the brains of aging employees and can soon be lost if not digitized.
Echoed by conversations at BuiltWorlds' AI/ML Inaugural Annual Meeting (February 2024), Buildings Conference (May 2024), and the Paris Summit (June 2024), data privacy is top of mind and something that each organization has to come to terms with. Increasingly, groups are deciding to build solutions internally using their own data, which presents a separate set of challenges from scaling and maintaining the solution to integrating with existing tech stacks and the number of resources required to develop the technology, which is why our panelists focused on knowing what the technology is for and how you want it to impact your business.
As long as BuiltWorlds has covered AI/ML, we have seen hesitation from the end-users. BuiltWorlds' own Sofie Richardson outlines this struggle in her article "To Share or Not to Share: The Construction Contractor’s Data Dilemma," pulling insight from Cutler Knupp (Haskell/Dysruptek) and Jay Cami (Construct AI). Click here to read more!
"Don't Trust, Always Verify"
AI/ML solutions are not foolproof and the output is often only as good as the input. The panel referenced that when using large language models (LLMs) for example, you can ask the same question several times and get vastly different answers. However, there are times when there is one, objectively correct answer that cannot be misinterpreted. As such, Dan Williamson is developing a company-wide curriculum so their employees can use the technology properly. He went on to mention that traceability and being able to verify where data is coming from is something that can build trust in the product internally.
Changing gears slightly, Ivy Nguyen mentioned that there was a prior wave of AI [in 2015] that was eerily similar to what AI vendors are pitching today. The bulk of the value is in “AI panic” to advance digitization, but the only ones that lasted are the companies that delivered real ROI and are continuing to deliver ROI.
The 2024 Construction Tech Conference underscored the transformative potential of AI/ML within the AEC industry, particularly highlighted by this panel of experts. As AI/ML continues to evolve, BuiltWorlds will be your one-stop shop for critical insights driven by objective research and high-level conversations for organizations looking to effectively integrate these technologies and harness their full potential.
Want to Stay in the Loop?
Contact research@builtworlds.com or sales@builtworlds.com for more information on the AI/ML Research Track and Forum Group which seeks to prepare our member companies for the potential of AI/ML, its applications in the AEC Sector and also the role of startups, cloud computing companies, sector-focused large tech companies, academia, and the industry in this period of change.
Discussion
Be the first to leave a comment.
You must be a member of the BuiltWorlds community to join the discussion.