A CEO and COO’s Perspective on Industrial AI Platforms

At its core, manufacturing has always been about productivity: turning input into output as efficiently, reliably, and cost-effectively as possible. But delivering on that promise has never been more difficult. Today’s manufacturers are expected to produce a wider variety of products, in more places, at higher quality, on tighter timelines—all while facing growing economic pressure and rising complexity across their operations.

“The ultimate goals of manufacturing haven’t really changed,” says Dr. Jan Michael Mrosik, former COO of Siemens Digital Industries. “But the level of expectations, professionalism, and mastery needed to stay competitive is going up and up and up.”

We caught up with Dr. Mrosik to get his take on what’s driving that shift. In this Q&A, he shares his perspective on the evolving role of data, the next S-curve in digital transformation, and why Industrial AI Platforms are quickly becoming a must-have for manufacturing leaders.

A COO’s perspective must align with the broader views and expectations of the company. That said, I believe there are five main goals every COO, CEO, and CFO would agree on: value creation, efficiency, quality, cost, and, increasingly, flexibility.

Fundamentally, manufacturing is about maximizing output while minimizing input. That means making the best use of your resources, ensuring consistent quality, and keeping costs low. Every part that needs to be reworked or scrapped is a waste of time and money. And cost, in the end, reflects how well you’re managing both efficiency and quality. Flexibility, meanwhile, is becoming a real differentiator. You need to handle different products, configurations, and volumes—often on the same line—without losing profitability.

It used to be that you could build one big factory somewhere and produce for the whole world, maximizing economies of scale and keeping things relatively simple. But a few years ago—starting even before COVID and then accelerated by it—we saw that model start to get disrupted. Today, more and more procurement and manufacturing is done in-country for in-country or region. The value chain is breaking up.

That creates real complexity. You now have to replicate production and procurement processes across multiple locations, often with different suppliers and varying levels of capability. And in complex manufacturing environments, as we all know, things tend to go wrong. You need know-how, you need capabilities, and you need tools to manage those problems systematically.

At the same time, the economic pressure has never been this intense. Western companies are being challenged by fast movers from other parts of the world. If you want to stay competitive, you have to squeeze productivity to the last level of detail. Where automation was enough a few years ago, now it’s time to climb the next S-curve.

If you look at the improvement journey in manufacturing, I’d say most companies that were going to master automation already have. The ones that haven’t probably never will. That was the last S-curve. The next one is all about connectivity and data.

Many companies have started collecting huge volumes of data. They’ve connected machines, added sensors, built data lakes. But now they’re asking, “What do I do with all this data?” It doesn’t help if it’s just sitting on a big hard drive somewhere. The next challenge is unlocking the potential of that data and using it in a systematic way to improve factory performance.

That’s easier said than done. Data alone doesn’t tell you much. You need to analyze it holistically, with the right tools and context. It takes consistent data, a systematic approach, and the right expertise to turn it into meaningful business and manufacturing intelligence.

Someone once told me the biggest problem sits between the screen and the backrest of the chair. I believe this is true in manufacturing too. The challenge isn’t technology. The tools are already there. Companies like EthonAI have solutions that are ready to use today. The problem is people who either don’t want to embrace, or haven’t yet understood, the value of data and what’s possible when you fully tap into it.

That’s not unique to manufacturing. We saw the same thing with electric vehicles, or the introduction of automation. Every time there’s a new S-curve, you have early adopters, then a following group, and then the hesitant ones who resist the change.

Know-how is another constraint, especially in smaller companies. Many of these companies don’t have data scientists on staff. And in some cases, you’ve got owners who’ve been doing things the same way for decades. Changing that mindset takes time.

First, you need the basics in place. Automation, connectivity, a big data system that reliably captures what’s happening on the shop floor. Whether you store that data in the cloud or on-premises doesn’t matter as much. What matters is that the data is there, consistent, and reliable. This is what I would call the necessary condition.

But collecting data isn’t enough. What actually makes the difference is having a system that uses it—analyzes it, interprets it, and gives clear, practical guidance when something goes off track. A good system continuously helps you optimize performance, whether that’s yield, bottlenecks, inventory, efficiency, and so on.

This is where AI comes in. It’s the most powerful and flexible tool we have for understanding what’s really happening in complex environments. The data might contain the answers, but without AI, those answers often stay hidden. AI gives you the ability to connect the dots, to pinpoint the root cause of a deviation, to act quickly. The speed of progress here is truly mind-blowing. What’s possible today already goes far beyond what we saw just a year ago. And it’s improving by the day.

The Industrial AI Platform is the perfect system for doing the real analytical work—taking your production data, interpreting it systematically, and turning it into practical, actionable guidance. It’s not just about dashboards or data visualization. An Industrial AI Platform helps you understand where and how to optimize across a wide range of manufacturing dimensions.

That could mean improving first pass yield, improving quality deficiencies, or understanding the root cause of a bottleneck. It might flag inefficiencies in how you’re using your equipment or show you where you’re carrying excess inventory. Whatever the case, an Industrial AI Platform takes the data you already have and uses it to improve factory performance, continuously and across the board.

Industry 4.0 and OT systems like MES laid the groundwork by automating deterministic processes and capturing data from the shop floor. But their primary role was execution: running the factory, collecting information, and ensuring traceability.

An Industrial AI Platform builds on that foundation and adds a new layer of intelligence, using causal AI to model your manufacturing process, correlate it with real-time data, and pinpoint issues that would be difficult—if not impossible—for a human to detect. It doesn’t just show you what’s happening. It helps you understand why it’s happening and what to do about it.

In that sense, an Industrial AI Platform is the logical next step after Industry 4.0. It’s about making sense of your data in a systematic, holistic way and turning it into real business results.

Get started. That’s the most important thing. The technology is mature enough to begin today, and it’s only getting more powerful and intelligent over time. It takes time and experience to use it well, so the earlier you begin, the better prepared you’ll be to capture the benefits.

I’ve seen too many leaders and companies analyze things to death—the complexity, the resources, the ROI. It’s right to have these discussions, but the important thing now is to get going. We have disruptive technology that has very, very high potential. Embrace the opportunities that an Industrial AI Platform offers in terms of continuously optimizing your operations. Don’t be the last to realize it. Start now, get on the journey, and refine as you go.