I'll be heading to Chicago from June 22 to 25 for Automate 2026 to see firsthand how the industrial automation landscape continues to evolve. The last time I attended the event (2024), many conversations centered on digital transformation, connectivity, and early Artificial Intelligence (AI) experimentation.
Two years later, the industry looks noticeably different. Physical AI, humanoid robotics, Software-Defined Automation (SDA), and increasingly intelligent industrial systems have transformed from immature concepts to Return on Investment (ROI)-generating deployments.
As I walk the show floor, I'll be looking closely at how these technologies are progressing, where adoption barriers remain, and which innovations are positioned to deliver meaningful operational impact. Ahead of the event, I sat down to reflect on five key questions that will shape the industrial automation story in 2026 and beyond.
What feels different about Automate 2026 compared to previous years?
I get the feeling the event itself is larger and more impactful. Automate is a great show because it brings together companies from across the entire industrial automation value chain. It's not just robots, machines, or software. It's the entire stack.
What I've observed in the past is that many of the booths are similar in size, which creates a more egalitarian environment. You don't have one or two companies dominating all the attention. That's important because automation, and robotics in particular, is characterized by a long tail of vendors rather than a handful of dominant players.
What's changing this year is the rise of Physical AI and AI more broadly. Historically, industrial automation has been reserved for highly repeatable tasks or situations where organizations wanted to remove people from dull, dirty, or dangerous jobs. AI is changing that equation by making automation accessible to a much broader set of use cases, customers, and industries.
Instead of automation being reserved for companies like Ford, General Motors, PepsiCo, or Procter & Gamble, AI is lowering barriers to entry. Smaller manufacturers can now automate tasks that previously weren't worth the effort or complexity. Technology vendors such as Palladyne AI, Siemens, ABB, Schneider Electric, Rockwell Automation, and NVIDIA are all helping push the industry in that direction.
What questions are you hoping Automate 2026 helps answer?
I'd love to get a better understanding of the current state of Vision-Language-Action (VLA) models as they apply to industrial environments. More broadly, I'm interested in seeing whether any companies are doing something truly revolutionary in the automation space. This means paying close attention to how industrial automation interfaces are evolving. How do users interact with these systems? How easy are they to implement on the shop floor or in the back office?
I'd also like to see more digital twin concepts embedded directly into automation frameworks. The opportunity isn't simply making automation easier to deploy, but making it easier to optimize. Manufacturers are constantly dealing with change management. Companies like Apple and Samsung release new products every 9 to 12 months, which means factories and supply chains are constantly adapting. Industrial automation systems must keep pace with those operational changes.
Another area I'm watching closely is automation oversight. Most manufacturers operate mixed environments that combine legacy infrastructure with new technologies from multiple vendors. It will be interesting to see how vendors will showcase interoperability rather than technologies operating in isolation.
How do you define Physical AI, and why is everyone talking about it?
Physical AI is often discussed in the context of robotics, but I think that's too narrow a definition. To me, Physical AI is really about automating automation.
In robotics environments, that might involve Autonomous Mobile Robots (AMRs), Collaborative Robots (cobots), or drones. But Physical AI is equally relevant in process industries where there may not be any robots at all.
Think about a valve, pump, heat exchanger, or other industrial asset. Physical AI ensures those systems automatically adapt based on sensor inputs and changing operating conditions. The intelligence sits at the edge and continuously adjusts how equipment operates.
Ultimately, the role of Physical AI is to close the loop between digital insights and physical actions. That's why I see it as a critical component of the future digital twin vision. The goal isn't simply gathering data. Instead, Physical AI should help create comprehensive systems that can respond to what that data are telling them in real time.
This is one reason NVIDIA's Physical AI initiatives have ignited immense buzz across the industrial sector. By combining simulation environments, foundation models, and robotics platforms, manufacturers can train and validate automation workflows before deploying them in production. Similarly, Palladyne AI is focused on making robotic systems more adaptive, helping automation tackle tasks that were previously too variable or unpredictable for traditional programming approaches. This is facilitated through the observe, learn, reason, and act closed-loop framework.
But to get to this point, manufacturers must eliminate data silos. Today, industrial organizations only use about 5% of all the operational data they generate. This makes data fabric innovations such as HighByte’s Intelligence Hub and Velotic’s Proficy essential tools for contextualizing data and enabling real-time action.
Why is software-defined automation becoming such an important trend?
The significance of Software-Defined Automation (SDA) is that the virtualization of control architectures creates more flexible, adaptable, and resilient automation workflows. Moreover, AI is increasingly integrated into SDA frameworks to help automate engineering and operational tasks.
Ten years ago, we focused on Machine-to-Machine (M2M) communication and basic visibility. Manufacturers wanted to know whether a machine was running, whether it was hot or cold, or whether a process was operating correctly.
The last decade has been about contextualizing that information through Internet of Things (IoT) platforms, analytics, and device management tools. Now the next phase is taking that information and enabling a level of control. Based on what we know about connected assets, what action should be taken?
Ideally, those actions are centrally managed. Operators shouldn't need to physically visit every machine whenever conditions change.
I often use the example of modern snowmaking systems. Years ago, workers had to climb mountains in the middle of the night to manually turn snow guns on and off. Today, those systems can be controlled remotely, monitored continuously, and adjusted from a central location.
The same thing is happening in industrial automation. We're moving away from managing equipment directly at the point of operation and toward centrally managed environments that allow operators to be far more effective with their time.
Siemens has been particularly vocal about SDA through its support for virtualizing industrial operations and simplifying automation deployment. Underpinned by the industrial edge, the German vendor empowers Operational Technology (OT) teams to conduct runtimes in a virtual software environment and operate in a more Information Technology (IT)-like fashion. Beckhoff's Personal Computer (PC)-based control architecture also reflects the industry's shift away from rigid hardware dependencies toward more flexible software-driven environments.
What comes next in industrial automation?
The industry is entering a period during which automation becomes more than an operational outcome. Several technological developments are converging to push manufacturers toward higher levels of autonomy, greater use of simulation, and more intelligent decision-making:
- Automation becomes an input, not just an output. Today, most industrial organizations treat automation as the end goal. Looking ahead, automation should become an input into closed-loop system design. That's how we'll achieve higher levels of autonomy.
- Closed-loop systems drive continuous improvement. The idea is to create systems that improve over time based on what is happening in the real world. As trust and determinism increase, more actions can occur automatically without human intervention. That changes the role of engineers and operators. Instead of creating processes from the ground up, they'll increasingly focus on approving actions, managing exceptions, and triaging issues.
- Digital twins and simulation become operational tools. Industrial automation naturally extends to digital twins and simulation. Companies are moving toward virtualized automation environments where they can simulate outcomes, optimize processes, and understand how changes affect not just a single machine or work cell, but an entire factory.
- Vendors are accelerating virtualized manufacturing. Siemens, Ansys, Dassault Systèmes, and a handful of other vendors are helping advance this transition by expanding simulation capabilities across product design, manufacturing operations, and factory planning. The long-term goal is to give organizations the ability to test changes virtually before introducing risk into live production environments.
- Simulation adoption still has significant room to grow. Every company producing physical products already relies on Computer-Aided Design (CAD) software. They couldn't function without it. However, industrial automation maturity and simulation adoption still vary widely across manufacturers. On the upside, barriers to entry are falling as a result of vendor innovation. It’s becoming much easier for smaller, talent-constrained manufacturers to generate ROI from simulation tools.
Let’s Connect at Automate 2026: If you’re attending the conference, let’s connect to discuss the biggest trends in industrial automation, share your experience, or brief me on your company’s latest innovations.
Carter Gordon