It’s not a question whether the artificial intelligence revolution will continue. It’s a matter of how AI trends will shape the future of the infrastructure industry. With it comes many questions about job security, data security and how to capitalize on this emerging technology while protecting your business interests.
Roads & Bridges’ panel, Getting Ready for AI: A Panel Discussion with Engineering and Technology Leaders, recently brought together experts from consulting engineering firms and software vendors to discuss the topic.
In the panel moderated by Jalpesh Patel, then business development manager of infrastructure for ALLPLAN (he now serves as Industry Development Manager – Infrastructure at Bluebeam), three experts explored the most pressing questions about AI in infrastructure and how they see it shaping the industry’s future.
Defining AI
Before diving into how AI will transform the infrastructure industry, Patel asked the panelists to define AI.
“AI, in a general sense, is about developing software or machines that have something that appears to be human-like intelligence or can do things that humans would typically be required to do in the software space,” said Terry Walters, the digital delivery evangelist at Maldelo and founder and chief architect at RoadCADdie.ai. “Essentially, it means making software that can learn in some fashion and then use that learning to solve new problems.”
While AI feels brand-new, its beginnings date to the 1950s, when several developers built applications that could learn how to play checkers, Walters explained. From the 1960s to the 2000s, AI’s growth was steady and flat until increasing in the 2010s until today, when it has become a part of most people’s lives.
“AI represents a new toolset and a new capability,” said Don Jacob, the chief innovation officer at Bluebeam. “We’ve recognized the application of the tool is important, but we are focusing on how we help people get projects done better, get the world built better, sustainably in cost and under schedule.”
Eduardo Lazzarotto, the chief product and strategy officer at ALLPLAN, sees AI as a co-pilot to assist and automate what people are doing—not replace them.
“One of the first questions that users and the industry in general have to ask themselves is how do they see the future of AI?” he said. “And what do they want that solution to deliver within their current workforce?”
Prioritizing security and teamwork
According to Walters, the explosion and proliferation of AI tools, especially free tools, are driving companies to shift the allocation of resources and capital.
“People are becoming more efficient because they’ve started to adopt these tools,” Walters said. “They’re either able to get more done or focus on the things that AI still isn’t good at. The important thing though is the security piece, especially when folks are going out to publicly available tools and [inputting] privately contained information.”
Experimenting with AI is the first step, but Jacob said the critical next step is “explainable AI,” which are tools and methods designed to help people understand the results of machine learning. Specifically, these tools are going to be essential with what Jacob calls mission-critical scenarios where human life is at stake.
“Being able to understand why the machine gave you the answer is going to be something very important for us,” he said.
Lazzarotto added that getting to this point is going to require teamwork and collaboration.
“We are always trying to integrate with other solutions to make sure the client has the workflow that they feel is right,” he said. “We’re not trying to force them into a certain aspect of using technology.”
Using AI in the AEC industry
When Walters worked on a recent Texas Department of Transportation project, he generated a “frequently asked questions” document with AI because he had a large amount of data from numerous stakeholders. AIenabled him to condense columns of information from a spreadsheet in 20 minutes instead of several hours. He has also seen AI used in grading, mechanical engineering and circuit design in electrical engineering.
“AI will be one of the most important technologies we have developed, and it will impact us in ways we can’t yet understand,” Walters said. “It’s kind of like trying to guess what the internet was going to become in the 1990s.”
In addition to consolidating text, Jacob sees significant opportunity with AI’s ability to synthesize information across different data types including text, semi-structured data, graphics, drawings, models, photos, video and audio.
“I think that is a real opportunity [to be able to apply] all the structured, semi-structured, and different types of data and see how it is being applied across other industries,” Jacob said. “I also want to underline that this is going to take all of us in the industry coming together as we go into this season of innovation with AI.”
As with every technology and tool that has come from AI, quality is essential. AI tools will only be as good as the information used to “train” it.
“Great AI comes from great data,” Lazzarotto said. “More than ever, we need to remind the whole AEC/O industry that today’s data is still locked within files, workflows or processes that need to be open.”
