Build Better Bridges by Creating a Culture of Innovation and Harnessing AI

Aetna Bridge, a general contractor based in Rhode Island, is no stranger to bridge construction. In the 70 years since their founding, they’ve worked on over 1,400 bridges across New England. A track record like that doesn’t come by accident, and in a lot of ways it’s thanks to the disciplined practice of means and methods that have been honed over decades — day after day, year after year.

In 2022, however, Aetna Bridge did make one key decision that would lead to big changes: the decision to adopt Procore and embark on a journey of digital transformation. Prior to that, Aetna Bridge ran on spreadsheets, paper files, and clipboards. But by the second half of 2023, the organization was radically transformed, having built a true culture of innovation and achieving results in their quality control program that had been previously unattainable.

Today, the company is at the forefront of construction technology, developing and testing a custom artificial intelligence (AI)-based quality control management tool. This tool gives them deep, actionable insights into their operations and opportunities for continued improvement.

But the journey didn’t begin with AI. Morgan Field, a project engineer at Aetna Bridge who has spearheaded the company’s digital transformation, insists that building a strong foundation is critical.

“We couldn’t have just jumped to having an AI quality control manager,” explained Field. “Every single step that we took to get there was fully necessary.”

A Story Behind Every Bolt

As a New England bridge builder, Aetna Bridge is, at its heart, a self-perform steel erection company. The American Institute of Steel Construction (AISC), is a key trade organization in the U.S. The AISC’s Erectors Certification is a requirement for any contractor that wants to perform bridge erection for Federally funded projects and most states require it as well.

“Holding this certification is something that’s extremely important to us,” said Field.

To maintain the certification, the AISC audits Aetna Bridge annually to make sure they are compliant with AISC’s quality, safety, and technical proficiency requirements. Prior to Aetna Bridge’s digital transformation, the audit was a high-stakes scramble to gather the proper documentation and make sure that records were in order. But that’s no longer the case.

A single structural connection in a steel truss bridge is made with dozens, even hundreds, of bolts. The AISC requires a bridge builder to document the entire history of every one of those bolts, from delivery, initial inspection, handling and storage, certificates of compliance to installation and finally testing. That’s because there’s no margin for error. When the bridge in question is carrying thousands of vehicles a day, there can be no tolerance for faulty materials or work.

A Perfect Use Case for Procore’s Action Plans and Inspections Tools

Aetna Bridge used Procore’s Action Plans tool to encapsulate the entire lifecycle of the AISC-compliant bolt in a single place. At each step in the bolt’s journey, the Action Plan includes step-by-step instructions, provides reference documentation, and links to the team’s standard inspection templates related to that step.

Field is quick to acknowledge that not every construction company is going to care about bolts, or even steel erection standards. But he emphasizes that no matter the type of construction or trade you’re involved in, there’s sure to be a critical material or field method that must be carefully controlled for quality and documented without fail — and so he sees Action Plans and Inspections in Procore as having extraordinarily broad application.

The benefits of centralizing everything about a bolt in one digital artifact is powerful on a number of levels. For one, Field describes how the Action Plan and the information they’ve linked to it help professionals in the field standardize their practices.

“No one ever has the excuse to not do it the right way, or not to know how to do it,” said Field, because everything is linked to one place.

Another benefit is the ease with which the information can be found or referred to in the future. Because everything about the bolt is centralized, it’s a simple matter to hunt down the information that’s needed in case of a question or issue.

“I can tell you, from the day it was manufactured to the day it was installed, everything you need to know about [any bolt] in a minute,” said Field.

But the most valuable benefit of this approach is the data itself.

High Quality Data, High Value Asset

“The importance of [capturing digital data] isn’t just easy access, it’s what it does for your company,” explained Field. The rigor that a standardized QC process brought, plus powerful and flexible reporting, led to an extraordinary outcome:

We got our first perfect AISC audit in the history of our company.”

– Morgan Field, Project Engineer at Aetna Bridge

Given that Aetna Bridge simply cannot operate without their AISC Erector’s Certification, the value of a successful audit — much less a perfect one — is inestimable.

A report like the one Aetna Bridge uses today to streamline their AISC audit, however, is not the end of what they’re doing with their data. Digital data is far more valuable than its paper-based equivalent because of all the ways you can manipulate it once you have it. That potential is even higher, Field explained, now that AI-based tools can unlock the potential of what he calls “unstructured” data, data as in meeting notes in natural-language text, or pictures.

Custom AI Tools to Leverage Structured and Unstructured Data

Seeing that potential, Field began developing custom AI tools to test what was possible.

“My main [use case] is a quality control inspector specifically for the AISC standards,” explained Field. It uses the data like that captured in the field via the Action Plans tool in Procore, as well as Time Sheets, Daily Logs, and more. Field has been able to achieve three important and remarkable results for his company using AI.

The Value of Good Summarization

One of the “most basic functions” of generative AI is to take a large data set and summarize it, according to Field. While that may be basic functionality, any project manager or field lead who has spent half a day sorting through reports looking for the one they need to resolve an issue on the jobsite, or who is trying to keep the job moving while getting up to speed on an enormous body of project documentation, can instantly recognize the value of a good summary.

Field was able to use AI to create four-to-five line summaries of the work completed in a week, boiling down hundreds of individual Timesheet entries, Daily Log entries, and other data into an easily digestible report that a company executive or a project owner could consume in a glance. He also was able to use AI to pick important insights out of vast amounts of data. For example, a single instance of noncompliance from reems of daily reporting. Simply surfacing important details that might have otherwise remained hidden by the large volume of data gave his company the opportunity to act and improve it.

Knowledge and Conformance to Standards

Things got even more interesting when Field used his custom AI tool to essentially “compare” project data to the AISC spec book.

By infusing the AI tool with the AISC Certified Erector Standards Field was able to ask the tool to surface the parts of the standards that were relevant to the work being performed.

“Our system sifts through all of our project data, which no one would have time to do, and the entire spec book, which no one would have time to do, and finds exactly what we should be looking at for the work to be done,” Field shared.

The potential time savings for field professionals in this application are enormous, because the AI can direct them to the portions of the spec book that they should review in connection with current scope of work. Even more important, is the risk mitigation achieved. Field stresses that in practical terms, because of the imperative to keep the job moving, field professionals often can’t take the time to research hundreds of pages of standards. This means that work proceeds without any reference to standards at all. The AI changes this by making it possible to check in with relevant passages in the spec book without slowing productivity on the jobsite.

Action Item Recommendations

Taking things one step further, Field showed that generative AI was capable of producing context sensitive, non-arbitrary recommendations about how the Aetna Bridge team could improve its operations on their projects.

“I really want to emphasize that the richness of these recommendations is entirely dependent on the data that you are putting in,” Field said. “AI technology will not mean anything to your company if you don’t have your data ready to work with it. The data you’re collecting is the fuel for the AI. The more data you collect, the more valuable the AI will become.”

Field doesn’t expect every construction company to go start writing code and creating their own custom AI tools like he has. But, “the technology is coming.”

The important thing is to get ready to leverage AI by investing in digitization of the business and data maturity now.