AI is driving the building of data centers at a record rate. Gas turbine and power equipment vendors are scrambling to keep up with demand. The more than 40,000 manufacturers, metal fabricators, welders, and related professionals who gathered at the annual FABTECH show in Chicago in September of 2025 are feeling the benefits.
NRL Hall of Fame running back Emmitt Smith even showed up for a keynote. In the face of the AI transformation, he preached the importance of coming to terms with change. He recounted stories from his storied career about how a former coach wanted to move him from quarterback to running back when he was a young kid, failed to make the first team in college (for a short while), and of his retirement from the NFL and starting a new career.
“It’s our attitude towards change that really limits our possibilities,” said Smith. “Change happens to all of us; you don’t know what’s on the other side. When facing change, you need to broaden your game and prepare for opportunities.”
Case in point: MEC, the top metal fabricator in the US, is experiencing rapid expansion into the data center and power markets. It recently acquired another large fabricator known as Accu-Fab that has a large footprint in data center construction and counts electrical/power vendors like Eaton, Schneider Electric, and ABB among its clientele.
“Every week I am on calls about large data center campus projects from hyperscalers and others,” said Jag Reddy, President of MEC. “We were previously sourcing parts and materials from China, but U.S. capacity is now key as well as the ability to produce at high capacity to meet demand.”
The AI boom has led to shortened lead times, too. Previously, ME might deal with a 24-month cycle from RFQ to production. Reddy says they now have a couple of months. The company is currently testing out various AI tools to speed its processes.
Dane Manufacturing of Wisconsin is another manufacturer specializing in end-to-end metal fabrication and assembly. Troy Berg, its Owner and CEO, said his company expanded from less than a dozen employees to 250, driven by work for data centers and the power industry.
“We are expecting a really big 2025 and for it to continue at least until 2028,” said Berg.
In response, his company has invested $10 million in buying new equipment and automation tools for robotics, welding, blanking, forming, and bending metal to meet demand form the data center and power markets.
Traditional Skills Still Matter
Speaker after speaker at FABTECH stressed that traditional skills still matter, but they must also prepare for AI and automation. This applies to the gas turbine manufacturing and servicing sector too.
“AI may be overhyped, but it is the most important innovation in decades,” said Scott Sawyer, Co-Founder and Chief Scientist at Paperless Parts. “Manufacturers need to understand that AI is a technology, not a solution. It takes industry expertise to transform it into an effective solution.”
He laid out use cases for AI in industry such as incorporating it into computer aided design/manufacturing (CAD/CAM), design iteration, augmented reality headsets, process control/optimization, and one way he uses it with his business – to identify request for quotes (RFQ) and their many attachments and specifications to organize them more efficiently.
“AI can help your front office surface objective information faster so you can make informed decisions,” said Sawyer.
Cobots in the Workplace
If the FABTECH show is anything to go by, cobots (collaborative robots) are coming in record numbers. Large gas turbine OEMs have already deployed them in quantity. Expect mid-tier manufacturers and smaller shops to add them to their production lines, too.
“Automation using cobots should never be done without understanding the technology, processes, and especially the people working on the shop floor,” said Denise Stafford, Business Development Manager at industrial robot manufacturer KUKA.
She related a story about how one firm wasted $90 million in an all robot/cobot/automated factory for tool production that had no human workers. It was shut down as a flop. Reason: robots didn’t fully understand the nuances of the manual human actions involved in tool manufacture. She used this example to make the point that grand visions from leaders, managers, and innovators must be balanced with input and involvement from the shop floor.
“Respect and value your shop floor people and make them co-owners of new technology like AI and cobots,” said Stafford. “Instead of being slaves to the robots, find ways how they can help existing workers to learn and to make their jobs easier and more enjoyable, such as taking over tedious, repetitive, or dangerous tasks.”
Enhance the Human
Craig Ulrich, CEO of industrial automation firm Acieta, believes the proper role of AI is to enhance humans, not replace them.
“The human plus automation and AI is the perfect combination,” he said. “People are still needed to handle complex decisions, troubleshooting, and problem solving.”
He recommended directing AI and automation at roles with a high turnover to reduce the hiring burden and to spend time distinguishing clearly those tasks that can be automated. As automation is not one-size-fits-all, he added, understand what you do currently and what you want to do. In most cases, working with an experienced integrator is preferable to off-the-shelf implementations due to the specialist knowledge required.
“Know your product and process deeply if you want to effectively implement automation and incorporate AI,” said Ulrich. “Not all manual tasks have the ability to be automated so you need to set realistic expectations.”
