Making Visual AI Standard Practice in Complex Manufacturing - with Brian Ton of Florida Crystals Corporation

**Daniel Faggella** (0:12)
Welcome everyone to the Emerge AI in Business Podcast. Today's guest is Brian Ton, Senior Laboratory Manager at Florida Crystals Corporation. Brian examines why visual AI deployments stall between pilot and production. Not because the technology breaks, but because organizations struggle to build the operational trust that keeps these tools running shift to shift. The conversation covers the conditions that separate embedded deployments from shelved ones, closing the gap between technical teams and subject matter experts, establishing a feedback loop that functions as quality control for the quality tool itself, and starting with small wins before scaling to site or enterprise level. Today's episode is sponsored by Roboflow. A quick note for our audience that the views and opinions expressed by Brian on today's program are his own and do not reflect those of Florida Crystals Corporation or its leadership. Do you sell AI products or services? Emerge gives you access through trusted content and real conversations. Learn how leading AI brands like Nvidia and Google Cloud work with Emerge to reach Fortune 500 AI buyers. Download our media kit at emerge.com/ad1. That's emerj.com/ad1.
Now the conversation with Brian Ton.
Brian, thank you for joining me in our studio today.

**Brian Ton** (1:53)
Hi, how's it going, Yolande? It's really great to be here with you.

**Daniel Faggella** (1:56)
Great. So we had such a brilliant conversation the other day that I decided we have to get this recorded.
So our listeners are really in for a treat, and I think this is really going to land with those executives that are in quality roles in complex manufacturing. And I'm going to get us into thinking that there's this assumption that all the big quality problems are very much visible. And that if something big is broken, it will just pop up and we'll notice it almost immediately. And I think that's not the reality of it. The more we dig into this, the more we see that the real problems aren't the things that we see at every shift. They're the ones that kind of became background noise, and the things that we've learned to work around, but so we don't have to really label it as a problem anymore. And I'm curious what that looks like from your seat. What are the quality gaps that tend to stay hidden the longest, and why do they stay that way?

**Brian Ton** (2:51)
Well, I'd like to believe that those of us with responsibilities and quality are doing good at our jobs, and we're always trying to sniff those invisible things out, right? But I think some of those invisible things tend to come more visible in the world that we're in right now. So, you know, in the world of AI, I think change management is one of the biggest universal challenges facing many businesses. With technology constantly evolving, you know, AI has kind of put the accelerator on everything. And so even across the years, you can see new technologies and companies are throwing new solutions to old problems, right? And if you look at the whole life cycle of technology, some new technology emerges, right? And then companies will move to adopt those new technologies over time. And then they exist within the ecosystem of that business and reach a sort of level of maturity. Now, this takes time, right? And I guess one thing that we can acknowledge is that these changes are happening faster than individual societies and businesses can keep up, right? So before we can go through that cycle, oftentimes we find ourselves reaching for that new, that shiny new object. And even with certain technologies within my domain, we have, I can use an example of something that we employ, it's near-infrared spectroscopy. And that technology came out in, I think, the 70s or the 80s. I might have even seen a paper where it was kind of being introduced in the 60s, but I was in a conference just a few days ago.
And here we are talking about how to apply that technology in new ways. So it takes a lot of time for these things to mature and companies having to go to the change management to implement these new solutions.
It's very tough, very, very tough. So if we think of, I guess I could say metaphorically speaking, imagine looking through a telescope and then the lens of that telescope is constantly changing and the landscape that you're looking at is constantly changing. So, I mean, with all that being said, I think it's a huge challenge just to continue keeping up with everything.

**Daniel Faggella** (5:06)
Yeah, I like the way you phrased it. And I think this is, it's cool that we can repurpose or not necessarily repurpose, but find new ways to use old solutions. But then, like you said, at the same time, so many things are changing and with the changes, we also see new things coming. And I feel like a lot of the challenges that we see in this space right now is because we still see a lot of manufacturing plants relying on manual inspection.