One of the things I enjoy most about hosting the IFS Voices of Industry series is that it gives me the opportunity to learn directly from people who are shaping what Industrial AI looks like in the real world.

IFS Voices of Industry is a video series designed to ground the conversation around Industrial AI in real experience. Each episode brings together people across IFS who have spent time in the field, on the factory floor, or inside businesses, and who now shape how AI is built and applied through IFS.ai.

For Episode 2, I was honored to sit down with Stephen Jeff-Watts, SVP of Asset and Service Applications R&D at IFS, to do exactly that. Together, we explored what Industrial AI actually means for asset management and service operations, and why the way we talk about it matters.

Moving past the noise around Industrial AI

What stayed with me from the conversation was not a product pitch or a roadmap update. It was a much more grounded question: What does Industrial AI actually solve for the people running complex operations, day to day?

There is a lot of noise around Industrial AI right now, and I think many industry leaders are still figuring out what to make of it. Steve put it in a way that cut through that noise for me:

“Go through your business process, find the things that are causing you drag, and let’s try and partner to see where AI can remove that drag for you.”

“Drag” is a useful way to think about value

Cost drag. Time drag. Revenue drag. Profitability drag. When you start looking at Industrial AI through that lens, it stops feeling abstract pretty quickly. And that kind of thinking is what separates organizations that are genuinely making progress with AI from those still circling around it.

A practical example: first time fix rates

A great example we talked through was first time fix rates. Steve made the point that if AI can surface a broader view of asset history and help technicians diagnose issues before they arrive on site, a 25% improvement in first time fix rates is a realistic outcome. The impact of that single metric ripples through an entire organization: higher technician utilization, better customer satisfaction, lower operational costs. The business case starts to write itself. 

Steve also brought a perspective I found really valuable, drawing on time earlier in his career building infrastructure in the field. As he put it: 

“Getting a proper view of what’s been done and what’s been installed on that site is invaluable to the teams that have to look after it throughout its life.” 

That idea, capturing through-life data from the moment an asset is commissioned (not trying to retrofit it later, or approximating it from memory), is still something many asset-intensive organizations struggle with. And without that foundation, even the most capable AI has nothing reliable to work with. This is where having the right technology partner matters, especially one that understands the reality on the ground.

I also think this is the part that can get lost when conversations about Industrial AI jump straight to capabilities: the models, the automation, the dashboards. In the field, the quality and continuity of the underlying data is still what determines whether any of it is actually useful. 

The outcome that matters most: safety 

And then we got to the outcome I found most compelling when we talked about what success really looks like for customers. Efficiency gains and profitability improvements are meaningful, absolutely. But the one that stayed with me was safety. Steve said: 

“I want a customer to be able to say: we’ve gone more than a year without any safety incidents, and your software played a key role in making our site safer.” 

Industrial AI can play an important role in protecting people. For me, that is a measure that truly matters.