Throughout my career, I’ve been focused on developing technologies that shape the future of enterprise connectivity. It’s what drives me, because each advancement unlocks new capabilities that help transform how companies—and industries—operate.
But something different is happening now—something bigger.
We’ve entered the age of Industrial Intelligence, where AI, high-performance processing, and advanced connectivity are converging to power the next wave of industrial transformation. Across manufacturing, logistics, energy, and other sectors, companies are rapidly deploying autonomous robots, predictive maintenance, and AI-driven analytics to enhance efficiency, optimize productivity, and extract more meaningful business insights.
The problem is that most industrial enterprise networks weren’t built for this new era. I’ve had numerous conversations with customers and partners, and a recurring theme has emerged: AI is evolving faster than networks can handle. Legacy infrastructure wasn’t designed for AI-driven operations, and businesses are now grappling with coverage gaps, latency issues, security risks, and reliability challenges. This is forcing an important realization: we can’t bring AI to industrial environments without rethinking how industrial enterprise networks are built.
The Industrial Digital Divide
These conversations often stem from a more fundamental concern—the industrial digital divide.
Companies across sectors have invested significantly in AI and connectivity for their corporate offices. Their knowledge workers have access to the latest AI-powered analytics tools running on high-speed wired and wireless networks. But walk onto the factory floor, a refinery, or a distribution center, and you’ll often see frontline workers still using pen and paper to track operational data. They conduct equipment inspections, log performance metrics, and track inventory by hand, then walk to a workstation to enter that information into a system – a reality far away from the hype of an automated and digital workplace. That lag leads to errors, delays, and missed opportunities for better decision-making.
Industrial Intelligence depends on every worker, every device, and every process being connected in real time.
AI Processing Is Moving to the Edge
Another trend placing new demands on enterprise connectivity is the rapid rise of AI inferencing at the edge.
Over the last few years, AI has relied on training massive models in the cloud, using parallel GPUs and massive datasets to refine algorithms and make predictions. This is where most of the industry’s focus and investment has been.
But that model is evolving. We’re entering an era where AI workloads will shift from the cloud to real-time inferencing at the edge. Instead of depending on cloud processing for every decision, new AI models will run locally on devices, sensors, and industrial machines, enabling split-second decision-making without sending data back and forth.
However, "the edge" is not a single location—it’s a distributed model that spans multiple layers of computing. A significant portion of AI processing will occur on the device itself—whether that’s a robot, a camera, or an industrial sensor—but another critical layer exists at the network edge, where more intensive AI workloads are processed closer to the source of data.
The link between these two compute layers—on-device AI and network-edge processing—is becoming a critical requirement for the new wireless edge. This is where connectivity plays a defining role. The ability to seamlessly and reliably move data between these compute layers will determine how effectively AI can operate in industrial environments. Without ultra-reliable, low-latency connections, AI applications will be constrained by bottlenecks that limit their ability to deliver real-time insights and automation.
According to industry data, more than two-thirds of all AI workloads will involve inferencing at the edge within the next four years. This means industrial AI applications will need real-time, ultra-reliable connectivity to function. And as we move closer to physical AI and more intelligent and autonomous devices, the requirements for network performance will only grow more complex.
This shift poses a significant challenge for enterprise networks designed for client-server applications, not AI-driven operations. As AI pushes computing to the edge, smart CIOs and IT leaders are asking, “How does this impact my network?” It’s the right question to ask, and it’s critical that we answer it soon.
Laying the Foundation for the Future
The era of Industrial Intelligence is here, and it demands a new model for enterprise networking—one built to support AI-powered automation, real-time decision-making, and next-generation industrial applications.
In my next post, I’ll lay out a new approach for architecting enterprise networks that will serve as the foundation for the industrial AI era.
This is a pivotal moment for the industry, and I look forward to continuing the conversation.
