OEMs plan great escape from their comfort zone

There has never been a more exciting, or more terrifying time to be an original equipment manufacturer (OEM). “First and foremost, the world is changing more rapidly than we’re willing to admit,” said Evan Kaiser, vice president, global industry accounts, OEM, at Rockwell Automation, who led the OEM-to-OEM Leader Summit during Automation Fair 2025 this week in Chicago. “If you sit still, you’re falling way behind,” he warned.

You can’t lean on the past, warned Kaiser. “You can lean on Rockwell Automation for innovations and the support of those innovations in your machine designs,” he assured. “Many machine builders are comfortable with last year’s design,” Kaiser added. “But your competition is coming around behind you with a new design. Anchoring on the past is actually a barrier for future success.”

The new landscape is where value will be derived, said Kaiser. “Yesterday’s strengths are not tomorrow’s differentiators,” he declared. “Be prepared. Get out of the comfort zone. If we can’t escape from the old, we’re never going to get to the new. If you live in the past, it’s hard to get to the future. Change is accelerating. New ideas mean new results.”

The intersection of SDA, AI and robotics

The trajectory of software-defined automation (SDA), artificial intelligence (AI) and integrated robotics has brought us to this exciting, yet terrifying, intersection. The opportunities lie in fusing these three technologies, said Dan DeYoung, vice president, general manager, production design and control, Rockwell Automation.

“We have a heritage of building on and enabling technologies by fusing them together,” said DeYoung, who cited the original Logix controller platform as an example, fusing a discrete PLC (programmable logic controller) with a motion controller. “We’re now going to accelerate that and bring more contemporary disciplines to it,” he noted. “All things in a PLC are only as good as the brains behind the keyboard. How do we graduate from that and bring AI to the table so it can adapt and learn in real time?”

One manifestation of this intersection is integrated robotics. The autonomous mobile robots (AMRs) developed by the company’s OTTO division include controllers that manage both safety and navigation, explained DeYoung. “It’s adaptive. The navigation software and the control system are working together. Integrated robotics is control and AI fused together,” he said.

Software-defined automation isn’t just about a virtual controller or soft controller, DeYoung clarified. “It’s about changing the landscape of how things are designed by deploying those things into new forms,” he said. “If you look at CompactLogix and ControlLogix, none of that is going away. We’re committed to those platforms, but we’re bringing new variants of Logix.”

Design, test and deploy

AI has worked its way into a variety of technology applications, including machine design and digital twins.

“For decades, we’ve iterated our machine designs,” said Julie Robinson, business director, design software, Rockwell Automation. “While our machines have evolved, the way we develop our code really hasn’t. A lot of us version our code manually. We’ve had months-long commissioning cycles. Machines are moving from set-it-and-forget-it to continuous improvement. The code needs to change.”

How you develop software today is not how you’re going to develop it tomorrow, said Robinson, who recommended a three-step software strategy of design, test and deploy.

“Think of it as a system from the start,” explained Robinson. “You’re designing the overall architecture first. Software is a team sport with people spread out all over the world.”

Collaborative design means shared libraries. “Modern design means creating true libraries and sharing that across all projects,” noted Robinson. “They need to be centrally managed and version-controlled. Every change and iteration needs to be tracked, so if I have five people working on the same machine, I don’t need to worry about someone writing over someone else’s code. You can use AI to comment your code. Delegate things to AI that you don’t like to do or want to do. Delegate grunt work to the AI assistant.”

Testing should be an ongoing process, Robinson recommended. “In modern software, there’s not that handover,” she said. “Everyone is testing. There’s no test engineer. When you find errors in the late stages of a project, it can become much more expensive to fix those errors. We want to minimize those costs. It means unit tests, functional tests and system tests. Test imperfect conditions; test failure modes, catastrophic modes and timing issues. The goal is to decrease risk throughout the duration of the project, not just at the end.”

When design and test are done properly, deployment should be the simple part, said Robinson. “The idea is there should be no surprises,” she noted. “This is the easy part. It should feel like a very simple transfer. We want to keep it production-ready. You can deploy to one or many.”

Digital twins

The role of the digital twin has expanded. It’s being used for demonstration, simulation and virtual commissioning, explained John Pritchard, business manager, Emulate3D, Rockwell Automation. Digital twins can be used for animating a concept, simulating the physics and connecting programmable automation controllers (PACs) to the models and testing code and HMIs.

“The journey is how and where we test the code that powers this,” said Pritchard. “Unlike in the past, when we made a change, we had to test it physically on-site. Now, we can make those tests in the comfort of the office.”

One major reservation Pritchard hears about the digital twin is it seems like more to do, more work. “Is the juice worth the squeeze?” he asked. “Is the return on the effort worth it? Physics-based digital twins can accelerate time to value across the manufacturing industry.”