ProveIt! 2026 Conference Recap — AI Startups Making Their Case

Episode Summary

The episode titled "ProveIt! 2026 Conference Recap — AI Startups Making Their Case" focuses on key insights and discussions from the Prove It conference held in Dallas, Texas. The guests include Dylan DuFresne from Abellara, a consulting firm specializing in operations technology (OT) space; Glenn Gardner and Dylan DuFresne as co-founders of Abellara; Sam Elzner from Litmus, with extensive experience in industrial automation and data systems; Brad Hafer, an advisor to Fuse and a veteran in B2B software and technology; and Craig Scott, founder and CEO of Fuse, with background in automotive manufacturing. These professionals shared their perspectives on integrating artificial intelligence (AI) into product lifecycle management (PLM), operational technology (OT), and maintenance technologies.

During the conference, several critical insights emerged. First, there was a strong emphasis on the importance of bridging the gap between PLM, OT, and maintenance systems to create more efficient and integrated workflows. Second, the discussion highlighted the value of hands-on workshops over polished presentations, suggesting that interactive sessions provide deeper understanding and engagement among attendees. Lastly, the guests proposed innovative ideas for future conferences, such as hosting a live build-a-thon where attendees could observe real-time problem-solving by industry experts.

For PLM and engineering professionals, the key takeaway is to embrace cross-functional collaboration and leverage AI-driven solutions to enhance operational efficiency and innovation in their respective industries. The insights from these leading practitioners underscore the necessity of integrating diverse technologies for a more seamless product lifecycle management process.

---

Full Transcript

Michael Finocchiaro

And we're live. This is Michael Finnecara on the AI Across the Product Lifecycle podcast. This is sort of a special episode because we just wrapped up ⁓ the Prove It, the second Prove It conference in Dallas, Texas last week. ⁓ I have guests from Conferenzos, ⁓ Abelara and Fuse, well, sponsors, as well as Sam from Litmus. ⁓ gave one of the awesome presentations at the conference. ⁓ Nice to see you guys. You guys want to go around the horn and introduce yourselves? Maybe you start with Dylan and Sam, Glen, Brad and Craig. That's how I see you guys on the screen. So go ahead, Dylan.

Dylan DuFresne

Yeah, I'm Dylan, ⁓ co-founder of Abellara. We're a consulting firm and sponsored the Prima Conference. Yeah, background in the OT space.

Glenn Gardner

you

Michael Finocchiaro

Bravo. Okay. Sam.

Sam - Litmus

Awesome. Well, I'm privileged to hang with you gentlemen. ⁓ am Sam Elzner. I've been ⁓ working in the industrial automation and data ⁓ system space for about 20 years. ⁓ Started in big data storage and ⁓ object-based storage systems, distributed file system stuff, and then jumped into industrial automation. All my geeky enterprise IT friends would always wonder why I just watch

Glenn Gardner

you

Sam - Litmus

how it's made at home while I'm building a Chromium OS for this customer but could not get my head out of physical digital systems. So I found Kefware, just fell in love. It was meant to be. And then worked there for, geez, 15 years. I did a bunch of stuff and want to keep doing things. And so I have come over to Litmus to do a bunch of really cool stuff. So glad to be here. Thanks for having me.

Glenn Gardner

Mm hmm. ⁓

Michael Finocchiaro

Okay. Awesome, Sam. How about you, Glenn?

Glenn Gardner

Hey, my name is Glenn Gardner. I'm proud to be the co-founder of Abellara along with Dylan, who's also on this call. I'm a mechanical engineer by education, been in manufacturing about 25 years now. And I just recently moved to the dark side and became a consultant. But before that, for multiple decades, I was an end user in the trenches.

Michael Finocchiaro

Oh no no no no no no no no

Glenn Gardner

And I'm a rare mechanical engineer that really enjoys finance too. So invariably my mission involves linking the CFO and the CEO's office all the way down to the front line of the plant floor. And I love doing it. Dylan and I are having a great time building Abellara and thank you for having us Fino.

Michael Finocchiaro

Awesome good ⁓

Sam - Litmus

What's easier, systems or people?

Craig

Okay.

Glenn Gardner

⁓ man, systems. Systems by 100x.

Sam - Litmus

Right, right, right. Kudos to you guys for running businesses, seriously.

Michael Finocchiaro

You You

Glenn Gardner

Thanks.

Michael Finocchiaro

Hey Brad, how about you?

Brad Hafer

Sure, Brad Hafer basically spent my entire career in sort of this B2B software and technology space, going all the way back to my first job out of college selling IBM ES400s with Mapex and VPEX ⁓ to mid-range manufacturers. Spent about seven years in management consulting, traveling the world, spent a lot of time on plant floors in Asia and Europe, North America. And then the last 25 years spent in most of the three-letter acronyms of of B2B software, everything from PLM, you know, where like we go back and overlap slightly in our Matrix One, Deso days, to procurement, supply chain, ERP, and of course, MES most recently. And presently, I'm working as an advisor to a number of companies, including Fuse.

Michael Finocchiaro

Yep.

Glenn Gardner

Thank

Michael Finocchiaro

Awesome. How about you, Craig?

Craig

Yep, Craig Scott, founder, CEO of Fuse. ⁓ I go back as an end user, manufacturing engineer in ⁓ automotive industry for about 15, 16 years. And then went ⁓ to Plex for a couple of years, moved on from there, started my own system integration company. And then that's kind of where we got to know Ignuctive and some of the other, know, Kepware back in the day.

Glenn Gardner

you you

Craig

And then that's where we started to develop the Fuse platform.

Glenn Gardner

you

Michael Finocchiaro

Awesome. So for those joining that may or may not know, Prove It, it's a conference on MES IoT. ⁓ And what's curious about it is that ⁓ the companies like ⁓ Fuse and Litmus are given ⁓ three factories to solve a problem on. So maybe we can start with that, Sam, like the...

Glenn Gardner

you

Michael Finocchiaro

You guys ended up solving all three, but you want to describe the three scenarios that you were supposed to address in order to present it, to prove it?

Sam - Litmus

Yeah, gosh. So to be very candid with you, my main focus was on that two and a half minute tirade he saw me go through on stage. so ⁓ I really let the, not let, but I really saw Dave and Hussein really take the lead in terms of the three steps and everything. So I think, or the three different areas to engage. But what I'd say about the opportunity that Litmus saw was to

Michael Finocchiaro

You

Glenn Gardner

Okay.

Sam - Litmus

was really the core value of what the platform does. It connects to things, it breaks out stuff it finds in the outside world, it's a sensible subcomponents that lets you recompose those in different data products and then lets you egress them. But it leans into that idea that that's been hard, but it's solved. And the harder part too is how to scale that, how to repeat, rinse and repeat. So the Litmus team has leaned in really early on to one to many type of ⁓ perspective as they add a

Michael Finocchiaro

you

Sam - Litmus

feature here, it goes into an API, it goes into the management layer. We have good contextual tools to distribute the configurations automatically across different connected entities. you know, small little decisions like the entities will reach out to the management layer. It's not a solicited connection from management layer to target device. And all of that matters for ⁓ topologies built and

Michael Finocchiaro

Okay. Thank

Glenn Gardner

Okay.

Sam - Litmus

secure data flow and all that. I think for us, was about ⁓ making sense of the Proovit UNS, trying to show value by contextualizing information in a different asset ontologies, use SM profile machine identifier data type, for example, to do some of those. And then really show off the flexibility of the platform that we've built ⁓ that allows you to rinse and repeat really quickly. So I hope that answers the question and not...

Glenn Gardner

Mm hmm. So ⁓

Michael Finocchiaro

⁓ wait.

Sam - Litmus

to corporate sales either.

Craig

It's good.

Michael Finocchiaro

Well, actually I probably should have posted it to Dylan. I want to just for the audience to know the three use cases. I remember there's a bottling plant. Cause you were the guy, you were the, you were coding that in your truck doing that. Yeah.

Sam - Litmus

yeah, no, Dylan. ⁓ go Dylan. You're much closer. Yeah.

Glenn Gardner

Yeah, Dylan was the guy up at 3 AM doing it.

Craig

this.

Sam - Litmus

Yeah,

Craig

Yeah

Sam - Litmus

exactly.

Dylan DuFresne

There were.

Sam - Litmus

The guy who put it all together. Yeah.

Michael Finocchiaro

So I just, just very briefly that like what were the three use cases they had to solve.

Dylan DuFresne

Yeah, so there were three different factories you could choose to use one or more of them. ⁓ One was a unified namespace for a glass manufacturer. call that Enterprise A. The other one was a multi-site bottling facility called Enterprise B. And the third one was a life science example with a bioreactor for Enterprise C. And then anybody could choose to use one or more of those. know some companies like Fuse here, and I believe Litmus did as well, chose to do multiple of those enterprises. We focused on Enterprise B and the multi-site scale, but the choice was to use one of those to show a problem you solved, how much it cost, and how long it took.

Michael Finocchiaro

That's right. Mm.

Dylan DuFresne

and then there was a last minute one thrown into the CESME profile and those are their requirements. Everything else is open from there.

Glenn Gardner

Thank

Michael Finocchiaro

Thanks, Dylan. So on day one, Glenn, we started with a keynote of Walker who's kind of set that up and then his younger son Jared actually did a demo for us, right? And then we had Jeff Winter. So maybe you can tell us a little bit about how it started on day one.

Glenn Gardner

Yeah, exactly. Absolutely, yeah. so ⁓ Walker is famous for throwing curve balls to all the vendors there. That's part of the whole gig. Like if you're really in the trenches, if you've actually done system integration work, then you know that the plant is gonna throw curve balls at you at every corner. And so Walker does that. He's famous for doing that to each of the vendors there. He did it to himself. So on 4 a.m. on the morning of his big keynote address, ⁓ he threw away his prior script and kind of rewrote everything that he was gonna say.

Michael Finocchiaro

Ha ha ha.

Glenn Gardner

I really appreciated it. I won't go into detail right now, because what he ended up doing was sharing something very personal. ⁓ If you're close to Walker, you're aware that his mother died under some really horrible circumstances when Walker was just 10 years old. I'll leave it at that because I don't know how public Walker wishes to make that story, but that's what he talked about as his opening keynote. Whereas you would think he would have talked about something in the depths of like, how do you connect to a server and how can you layer AI on top of this particular thing?

Michael Finocchiaro

Okay.

Glenn Gardner

He knows all of that stuff, but I admired that. thought it was, it set the right tone for the entire conference because after hearing his deeply personal story and a deeply tragic story, you kind of felt this, like ⁓ you felt the community gelling. You felt like you were a part of something special. And for those, for while he was speaking, I was really less concerned about digits and bytes and machine learning and connectivity and PLCs and... You could feel it in the room. I admired the vibe that he set up with that opening speech.

Michael Finocchiaro

And then Jeff Winter did a really wonderful presentation on AI and how that was going,

Glenn Gardner

Exactly. Yeah. And if you're, if you're new to the community, ⁓ Jeff and Walker is like a super interesting one, two combo. Walker is famously bombastic. He tells it like it is. He loves controversy. He leans into all of that. And Jeff is the opposite. Jeff is very polished, very corporate. He's well-dressed, well-groomed. He sounds professional. And, ⁓ however, Jeff knows his stuff. He's been in the industry for over 20 years. He's worked at many different automation companies from many different angles. And so while Walker typically tells it,

Sam - Litmus

Yeah. Yeah.

Craig

Thank

Sam - Litmus

Yep.

Glenn Gardner

the story from the trenches, Jeff tends to be more top down. Jeff will tell the story from a macro perspective. He'll talk about big industry trends in a language that like a CEO or a CFO would be more easily able to absorb. And so Jeff's talk was great. I talked with Jeff Winter at the airport after the show was all over. and said, hey man, did you take, like, you really leveled up your game. Did you take professional speaking lessons or something that was, you've always been polished, but man, you took it to the next level. And he said, no, but I prepared for 60 hours. Jeff Winter puts 60 hours of prep into his talk and that's how it came off so polished. But yeah, getting, yeah.

Michael Finocchiaro

Yeah, it was really cool. Go ahead, sorry.

Craig

Cheers.

Glenn Gardner

getting into his content, there was one slide that he had newly created that was the most famous where he showed the connectivity between 10 different software systems. ⁓ And they were all of the systems that Brad listed earlier when Brad was talking about his experience. It was like PLM, MES, SCADA, supply chain planning, ERP, et cetera. He had 10 or 11 of those. And the graphic was intended to be suggestive, not like literally, this is the connectivity that you want to have between those.

Michael Finocchiaro

or. ⁓

Glenn Gardner

But he told the story of how each of these systems speaks a different language. They all need to communicate in an ensemble somehow. And that's the tough work that all of us at the community are doing. That landed really well. I heard a lot of follow up about that one slide in particular. And it's a really interesting combo to hear Walker speak and then Jeff speak. It's a great way to kick off a conference like this.

Michael Finocchiaro

Yeah, And it was a nice distraction from the HarmonyCon that was happening and ending as we were starting the.

Craig

Good contrast.

Glenn Gardner

Yeah.

Craig

⁓

Glenn Gardner

Thanks.

Craig

boy.

Glenn Gardner

I choose not to comment on that one.

Michael Finocchiaro

Well, ⁓ I sat on the litmus side after they split the room and I saw Machine Metrics and Tulip to give some pretty awesome presentations. I don't know if anybody was on the other side of the room. I think it was the ignition side. I don't recall who was speaking over there. You guys remember? Who was on that side?

Glenn Gardner

Yeah. ⁓ no, I had machine metrics and tulip and who was on the li- I don't remember who was on the litmus side, but-

Michael Finocchiaro

Yeah, I was trying to remember who was on the other side.

Dylan DuFresne

I day one, I've got it scheduled on a different monitor here. Day one, had on the inductive stage is Opto 22 software toolbox in Aviva. On the litmus stage was tool-it machine metrics.

Craig

Hmm.

Glenn Gardner

that's right.

Michael Finocchiaro

I heard the Avivo thing was amazing. I heard amazing feedback. yeah. ⁓

Glenn Gardner

Yeah.

Dylan DuFresne

Yep, and concept apply.

Brad Hafer

I feel 22 is interesting as well for a hard-hitting

Glenn Gardner

Viva.

Brad Hafer

solution. They really kind of nailed the problem.

Michael Finocchiaro

Really? Awesome.

Glenn Gardner

Yeah, you just listed five, five cornerstone vendors there. We should tear through them as quickly as possible, but they're all good ones. Fina, you want me take?

Dylan DuFresne

Thank

Michael Finocchiaro

Yeah, well, I interviewed ⁓ I interviewed Tulip and machine metrics on on site actually that that same on day three day two and they were really interesting. ⁓

Glenn Gardner

Nice. Yeah, we need a new category for those two. There isn't ⁓ ostensibly their MES, they're closest to MES, but that's not how I think of them. There's something different. ⁓

Sam - Litmus

Like the machine metrics folks? you could say like, ⁓ I would say machine metrics equals site machine, right? Like if we're doing analogies, right? ⁓ You know, get like, we run into them a lot with embedded ⁓ integration with like standalone equipment, CNC stuff, stuff that's not part of typical, ⁓ you know, production automation systems and are just sort of

Glenn Gardner

Machine metrics and tool up, yeah.

Michael Finocchiaro

Yeah, I agree. Yeah.

Glenn Gardner

Okay, yeah. Exactly.

Craig

Mm-hmm.

Sam - Litmus

for last 15 years to sort of the, some of the juiciest targets for integration because the OPC interfaces aren't there. The stuff had never been integrated to XYZ before. And so ⁓ I would put Tulip in like a category, ⁓ geez, mean, far less contextual, like Craig Fuse, right? Far less contextual than Fuse. More like, I mean, remember where Tulip started, right? They were started as an MIT Media Lab AR company.

Craig

Okay.

Michael Finocchiaro

Mm.

Sam - Litmus

right, broadcasting flow control directional arrows on floors from like embedded gadgets in the ceilings of aircraft terminals or airplane terminals. So like they have a really different take on tech even. It's almost, I view it as almost like a, it's a different common denominator that they found, but the interfaces are so composable that now they're layering in different contextual journeys for their.

Glenn Gardner

Mm-hmm.

Michael Finocchiaro

Thank you.

Sam - Litmus

customers, right? And they overlap with manufacturing quite a bit, HMI panels quite a bit. Like Craig, I think it's probably true of Fuse too. Maybe you can doodle on this a little bit, but like, just like Litmus written in a modern language, not necessarily going back to, you know, C++ for everything. So there's some efficiencies that Tulip is able to get, especially on the edge. It's like they're able to use languages that otherwise you'd be using C and just really difficult. So super high speed, high performance stuff at the edge, but

Michael Finocchiaro

Thank you. Thank

Craig

Mm-hmm.

Michael Finocchiaro

Right.

Sam - Litmus

ultimately like a bunch of molecules you can compose into your own anything, you know, ⁓ and I don't know if that I don't know Craig I mean to me ⁓ I almost look at tulip and fuse is complementary but but also maybe similar common denominator not enough from over my skis here now

Glenn Gardner

Yeah.

Michael Finocchiaro

Okay.

Craig

Yeah, no, think there's a, there's definitely some similarities there. You know, we kind of, we kind of talk about our approaches are quite a bit different. ⁓ Tulip is what I would say right to left in terms of building an application. You know, they start user interface and kind of work through and we actually go left to right. So we start with defining the models and then building the application around, you know, the, structure. So just different approaches. ⁓ you know, yeah.

Glenn Gardner

Yeah.

Michael Finocchiaro

Mm.

Sam - Litmus

Yep. Both really valuable, right? And there, for example, if someone has XYZ interface because of the right side angle that they came in, they might need models that Fuse is really good at or to tie into an application stack that Fuse is really good at. And sure, there's visuals and benefits of using Fuse, but it's nice to see the, you know, pluggable, chooseable, know, composable with third party tools combined together. Because that's been a...

Glenn Gardner

Yeah. Yeah. Okay.

Craig

Yeah, for sure.

Sam - Litmus

the way that many things in this space have come.

Michael Finocchiaro

Brad, you were mentioning OP 22. You want to say anything about what you saw there?

Brad Hafer

Well, was actually the first session I sat in on, so it was like all new

Glenn Gardner

Okay. Mm-hmm.

Brad Hafer

to me. And it began the week with discovery. And it was surprising, given that, sort of coming from that hardware side. I think to your point about Glenn, the different categories, there was 51 vendors there in attendance. Not all of them got on the main stage. I think that some of the folks should understand. So we didn't see everybody's. But there's a couple of different shapes and forms of those. ⁓ We talk about the stack, the connect, collect, store.

Michael Finocchiaro

That's right. Right.

Brad Hafer

analyze, visualize, you know, that Walker's famous for. I call it the water column because there's a lot of vendors out of the 51 that are in that, that water column or stack. But to your point earlier, there's these other categories that are more horizontal, like a machine metrics or a critical manufacturing or maintain X, you know, that really have application layers, not so much that OT up data stack. So it was really overall, I mean, it started with day one was just the six that you mentioned. It was a sampling of all different kinds of

Glenn Gardner

you

Brad Hafer

technologies and applications and the way they approach the problem.

Glenn Gardner

Yeah. Yeah. Yeah.

Michael Finocchiaro

I thought it got us to a strong start. ⁓ when we got to Tuesday, ⁓ then we had the presentation by Litmus you guys opened, and that was mind-blowing. Then the ignition one that just blew my socks off when they just deployed the whole factory in three minutes. Wow, just blew my mind, man. It was insane. ⁓

Sam - Litmus

Yeah. Yeah. Yeah.

Michael Finocchiaro

I thought we were really off and running by then. mean, we were already at a, you know, almost a sprint pace, right?

Glenn Gardner

Gotta let Sam talk about the litmus one. Yeah, that was my favorite of day two, man. I was pretty juiced after that one.

Sam - Litmus

I think that's, geez. Well, no, I'm just going to talk generally. I, didn't. Well, thanks. Thanks. Yeah. It was pretty great. I'm really proud of the team and they, the use case for agentic AI helping you troubleshoot is so key. know I would have saved a lot of time. And, know, when I do development often, if I'm developing applications that use ethernet as an IPC, I'm absolutely, you know, dumping packets into my query window and, and.

Craig

Thank

Michael Finocchiaro

You

Sam - Litmus

having a little guided session too. So it's cool to see more and more use of the kind of low level bits go to the AI. ⁓ One thing that would be really thrilling to me, ⁓ whether it's in a month, in a day, or even next year, doing that entire stage presentation again, just using a 24 gigabyte model running locally on a MacBook or a graphics card.

Glenn Gardner

Nice. You

Sam - Litmus

I don't think we needed a trillion parameters to figure out that someone had fat fingered a Modbus exception code, especially, or a Modbus port, especially when you put all the data in the query window ahead of time.

Craig

.

Sam - Litmus

just didn't. We have a few other things going on too. So we didn't have time to take that risk and iterate on it, ⁓ we have inside of Libis, we've actually purchased a couple of DGX boxes because we hope that we'll be starting to see these out.

Craig

it.

Sam - Litmus

the field a bit more and being able to run a 96 gigabyte model ⁓ or a 70 gig model and have 20 gig headroom for context window like this is making some meaningful progress it would just be nice you don't have to pay four thousand dollars for a golden box that doesn't even fit in a 42 urac yet you know so it's still early

Glenn Gardner

Okay. Okay.

Michael Finocchiaro

Well, probably within a year,

Craig

Mm-hmm.

Michael Finocchiaro

it'll all run just on a raspberry pie. The way things are going is so fast, right? It's amazing.

Sam - Litmus

You know, I'm not, I'm, you know, if you can, maybe a cluster of pies, Michael, right? Like low latency protocol between these pies, the brains like swapping things. And suddenly you've made a mouse brain with, with LLMs.

Michael Finocchiaro

Yeah.

Craig

I have a pie cluster. ⁓

Michael Finocchiaro

But Rick. It reminds me of Rick and Morty. Yeah, I've been working on that for a while. What really amazed me too and what impressed me the most with those two presentations is that I come from more of the PLM side where

Sam - Litmus

yeah, yeah, yeah, yeah.

Glenn Gardner

That was a solid impression. You even had the voice crack, Fino. That was incredible.

Sam - Litmus

That was nice. That was nice.

Craig

You

Michael Finocchiaro

Okay. We're doing AI, but it's more generative design and stuff. It hasn't really hit the infrastructure yet. We still are in relatively state of the art 2009 infrastructure with kernels as Brad knows that come from the late nineties. Right. I mean, we haven't really redone it. You guys are already coming in with freaking Kubernetes and Helm and YAML script. And I was just like, holy cow, that is just amazing how you guys have been able to take.

Glenn Gardner

you So you

Michael Finocchiaro

Infrastructure is code and which is an IT concept and apply that to OT. I that was really a brilliant move.

Sam - Litmus

Yeah. Thank you. Thank you. It's really nice to be at a place that had embraced that from the very beginning. I started to see my Kepler customers over the years, traditional Kepler customers. We've got GE Historian and Esqueda. ⁓ And maybe they're unique circumstances that have them be a little more of the ⁓ big data tech aware, CI, CD aware, but you know, tying into the REST APIs of these OT systems, cataloging their configurations and get using Visual Studio to program and deploy. when I started to see my OT customers do that, I desperately needed to give them a better surface area than just a 10 year old REST API, you know, ⁓ getting anything more done, just it, you can see the tech debt of some of the.

Michael Finocchiaro

Okay.

Sam - Litmus

because you detect it in real time when you have someone who didn't get the start in the 2015 paradigm, and it's bolting on. it's especially cool to be amongst Abelara and Craig because you guys are ⁓ starting with a fresh set of tools and some of these things are just a given and you didn't have to, they're almost freebies from doing it just differently.

Glenn Gardner

Yeah.

Craig

Yep.

Glenn Gardner

Yeah.

Craig

Yeah.

Michael Finocchiaro

Absolutely.

Glenn Gardner

And you make our lives a lot easier. Yeah, was after, Sam, it was after your presentation that the Fuse Ignition litmus stack really cemented in my mind. So yeah, especially Fuse and Litmus. Ignition, I mean, they're an OG. Everybody respects the heck out of Ignition, including me. They've been doing this for two decades. you know, major, major credit to Ignition, of course. But Fuse and Litmus.

Sam - Litmus

I'm so glad. Yeah, that was my

Michael Finocchiaro

What?

Sam - Litmus

Yeah, you've done, yeah, exactly. yeah, yeah. Yeah. Yeah. I mean, in my mind, we should be sharing our objects, Greg. Like we should just have the same definition of object at some point. yeah.

Craig

Yeah. What?

Glenn Gardner

Yeah

Michael Finocchiaro

Ha

Glenn Gardner

Yeah, but

Craig

I think the next song needs to be Fill the Gap, right? Yeah. Yeah.

Glenn Gardner

you guys attacked

Michael Finocchiaro

ha Ha ⁓

Glenn Gardner

there you go. Fuse ignition litmus. I love it. But seriously, like you guys

Sam - Litmus

Yeah, right. like that.

Dylan DuFresne

Yeah.

Glenn Gardner

attacked the same problem. If you attack the problem of, listen, I want to unite all of the OT data and you attacked it like frontline up or bottom up, if you will, you would build litmus and you would build it to scale. And that happened just in the last few years. If you attacked it from the enterprise down and you said, Hey, my starting point is I have PLM and SCP and ERP and MES and WMS and like, my gosh, all of them need to talk. You would build fuse.

Sam - Litmus

Yep. Mm-hmm.

Glenn Gardner

And all of the sudden on day two, this was always somewhere in my mind, but I was like, my gosh, that's the stack of the future. Fuse top down, litmus bottom up and ignition conduct tape absolutely anything between the two. All of a sudden I had a stack and I'm like, exactly. Yeah, you need that.

Sam - Litmus

Yeah.

Craig

you

Sam - Litmus

Yep. Exactly. Yeah, somebody's got to still do HMI and SCADA, right? Somebody still needs to do

Dylan DuFresne

Thank you.

Craig

Yeah.

Michael Finocchiaro

Yeah.

Sam - Litmus

that stuff. Ignition might have more competitors now too, right? Like that is such a hard thing to do. Real-time systems, 100 millisecond display, 5,000 variables, totally composable. Like they're very mature. But I remember in the Discord community, the I4O community, mean, a year ago, two years ago, there was...

Michael Finocchiaro

Mm.

Glenn Gardner

Yeah.

Sam - Litmus

Which is funny. I it was new when I got to IOT in 2011, ⁓ but still not that new. was like five years in, and now it's like 15 years in, and you can see the rough edges. You can see they had started when Java was the best way to do cross-platform, and they've had tech debt. It's neat to think that there might be another set of people who are interested in doing stuff like that. ⁓

Glenn Gardner

Yeah.

Sam - Litmus

And I don't think inductive would get mad at me saying this because we're all capitalists, right? Like we enjoy competition. We hope for that. It's going to make us all better and may the best one win, right? But it's cool to think that maybe there'll

Michael Finocchiaro

But.

Sam - Litmus

be other kind of mainstay players in the next 10 years and we'll all be complaining hard about inductive being too monolith and slow and all this,

Craig

Thank you.

Glenn Gardner

You

Brad Hafer

But to that point, Glenn, and not to jump ahead to Wednesday, but one of my favorite sessions of the week was what Dylan and Walker did and whiteboarded the solution. And I don't know if there's any preview and conversation, Dylan, with you and Walker beforehand, but just the architecture that you put up on stage was a selection of a very, very few number of the vendors out of the 51 that were present. And to that point of the litmus ignition and then ultimately Dylan adding the

Michael Finocchiaro

The Okay.

Glenn Gardner

Nope.

Brad Hafer

fused from the top down with a few other appendages was a great way to look at it.

Michael Finocchiaro

Yeah.

Dylan DuFresne

going into day three there a little bit, because it jumps right off of this fuse and litmus stack that we're talking about, is in that session, part of what we talked about was this minimum viable stack. And the reason, over the last year of workshops with various end users, call them, manufacturers, process facilities, across basically every vertical.

Glenn Gardner

Okay.

Michael Finocchiaro

Okay.

Dylan DuFresne

A common thread kind of came up, especially when we're talking ground field. We already have HMIs, already have SCADA, we already have something visualizing locally. But it's never, it's not usually connected. There's not always networks. It's oftentimes isolated to individual cells and machines and silos and various forms. And that's where Litmus can come in and connect to either existing OPC servers or devices or robots or everything else that could be on the plant floor at any capacity, including historians, file systems,

Sam - Litmus

file systems, databases,

Dylan DuFresne

and everything.

Sam - Litmus

you know, these blessed folks wrote a profinet, an asynchronous profinet interface for one of their embedded customers. Okay, cool. Yeah, anyway, literally anything, yeah.

Glenn Gardner

Awesome.

Dylan DuFresne

Yeah,

Michael Finocchiaro

You

Dylan DuFresne

and like anything on the plant floor without disrupting what's already there for visualization and controls. And if we need a SCADA layer, if we need something, then that's where ignition comes in. But brownfield, sometimes it's there, sometimes it's not. need to respect that. And litmus allows us to respect that.

Craig

Mm-hmm.

Michael Finocchiaro

you you

Dylan DuFresne

And then from the enterprise level, it's kind of the same story, kind of what Glenn was talking about from the different angles is the assistance have all the three letter acronyms. You got your PLM, you got your ERP, you got your supply chain planning, you got your warehouse management and all the other ones that I could spend an hour listening and still miss some. And Fuse allows us to connect to all of those at minimum in a very traditional iPad sense, but as a whole, it also lets us build those and respect what's there and fill the gaps that aren't. And so

Glenn Gardner

Yeah.

Dylan DuFresne

both Litmus and Fuse allow for that. And then things like Litmus, sorry, things like Ignition, things like Flow Software, things like High Byte, things like the MQTT brokers and all these other vendors are really good things that can also sit alongside that to fill individual needs. But the universal...

Michael Finocchiaro

Okay. you

Dylan DuFresne

near and not it's not 100 % but the near universal stack I've come to is we need something

Michael Finocchiaro

⁓ huh. ⁓

Dylan DuFresne

to connect to the plant floor when it's like connect the enterprise and the two tools that do that really undisputed on the market right now and in the name of capitalism at the expense of people on this call I'm actually hoping more come out but right now they're not there.

Glenn Gardner

Okay.

Sam - Litmus

Yeah, yeah, me too. I mean, was glad to have, at Kepler, was glad to have somebody else cater about drivers, because nobody else did, you know?

Dylan DuFresne

You know, so like there's

Glenn Gardner

Mm-hmm.

Craig

Yeah.

Dylan DuFresne

nobody who cares about drivers and is modern in the way that litmus does. And there's nobody that cares about enterprise connectivity and is modern in the way that fuses.

Sam - Litmus

Yeah. Yeah, no.

Glenn Gardner

Yeah.

Sam - Litmus

Yep, exactly.

Glenn Gardner

Amen.

Michael Finocchiaro

Well, I also wanted to say ⁓ there were some claims that made on the main stage that the vendors were kind of taken back, like Datto. There was a certain number that was thrown up there and the Datto guy...

Sam - Litmus

Yep. yeah, let's do this. What were the claims, Michael? Can

Glenn Gardner

No.

Sam - Litmus

you repeat the claims that we can get in trouble and reinforce these? No, but... ⁓

Michael Finocchiaro

There was a claim about

Craig

Yeah.

Michael Finocchiaro

response time that the data sky was like, ⁓ I'm a little worried about that. It's not exactly something like sub five millisecond or something like that. And he was like, no, we can do like 10. We can't get under.

Sam - Litmus

What was the piece, the response time? Oh, Walker had said you can do five. Yeah, I remember last year. I remember I had gone to prove it last year with Abby on my the general manager cap at the time she and I were were trying to drive drive stuff for a bit. And I remember Vatsalip on stage last year. And I think the message count per second came up and they were like, Vatsal, how quickly can you do that? And he was like, 100 milliseconds. And it's.

Michael Finocchiaro

something like that. You

Sam - Litmus

like 100 times faster than that. like the guy can't get up there and say, say what it's really capable of because somebody is going to hold them accountable to it. Right. So yeah, I have no doubt that Dados can probably do five. Right. Walker's probably right.

Glenn Gardner

Yeah.

Michael Finocchiaro

If.

Dylan DuFresne

and for For the record,

Michael Finocchiaro

Okay.

Dylan DuFresne

per their marketing on their public website, say status delivers sub 50 milliseconds end-to-end state readiness. And 2.89 medium latency. So those are the public.

Sam - Litmus

Ooh, even slower. Yeah, very good.

Glenn Gardner

Yeah.

Sam - Litmus

Yep. Yep. We always, we always use, we always use real time and then you could define whatever that means, you know, not a real time OS.

Glenn Gardner

Yeah. mean, if you use a...

Michael Finocchiaro

But

Dylan DuFresne

Yep. Yeah, so that's what's on their website as of today.

Glenn Gardner

Yeah.

Michael Finocchiaro

I just wanted to point out that I had the opportunity starting on day two and into a little day three, I did interviews of a lot of the people and they were really awesome interviews. So please go on my LinkedIn feed and see the interviews with ⁓ I had Tulip and I had Litmus and Ignition, sorry, Inductive Automation and ⁓ it was just awesome. It was amazing to learn from these people. And on the day three, interviewed Arlen and Thomas, the founders of the creators of MQT and OPC UA, which was just really, really fantastic. So again, I wanted to, yes, I did. And I want to just.

Glenn Gardner

Thanks.

Sam - Litmus

Oh, you talked to Tom Burke? Oh, that's great. Yeah. Yeah. think Tom, Tom has a new position now too, right? I think he, um, yep. Yep. Is he with inductive now? Yep. That's great. Yep. I think it's a good move. They've, they've leaned into UA for a long time in the way that Tom Burke really envisioned UA. You know, there, there's a lot to UA and it is a lot of things now that it always wanted to be. Like if you go back and look in 2006, they wrote it out.

Michael Finocchiaro

Yes. Yep. He moved across. Yes. He's sort of an evangelist for them now.

Glenn Gardner

Thanks.

Sam - Litmus

Someone like the exactly there's nothing new. It's not marketing literally like they wanted this map the world ⁓ class system that can model it out of model the universe be a graph. But the implementations looked a lot like process control and OPCDA and Tom is just that through and through. He's really

Glenn Gardner

Thank you.

Sam - Litmus

astute with object oriented interfaces and industrial automation. I really hope that a inductive can use his talents and promote promote that.

Glenn Gardner

Nice. That'd be awesome. Hey, Fino, can I ask Sam a sidebar question real quick? One of the takeaways that I had is, SESME and OPCDA and more modern UA, they're more tightly linked than I realized. But I admit this is not my area of expertise. Maybe Sam or Dylan could ⁓ wax eloquent on that topic.

Michael Finocchiaro

So let's. Sure.

Glenn Gardner

I had thought of Sesame Profiles as sort of starting from scratch, but to their great credit, there's a really strong OPC UA history there that maybe you guys can get into.

Sam - Litmus

Yeah. Yeah, the OPC Foundation released the first specification for unified architecture in 2006. And when they released that, they built on the OPC complex data, the classic old school object oriented interfaces are actually really useful. And like, let's maybe let people make their own instead of locking everything up into like OPCDA groups and tags, you know.

Michael Finocchiaro

Yeah. Okay.

Sam - Litmus

so tempting to be able to take that object that I can call functions against and then like make it Sam's data type. And Sam's function is going to take arm and drink coffee. Yay. Great. That's the hope, right? ⁓ So UA had been out there forever with a really detailed class modeling system with a mapping to a wire protocol, UA binary, UA, HTTP with XML and SOAP and all the 2006 trappings of modern tech, right? ⁓ And it know, it was, good. It's really good. But when you model the universe, you have to learn how to model everything from an atom to the universe, because that's the capabilities of the modeling system. And, you know, the software vendors who are building interfaces can do that because they're software engineers, but the average human ended up with stuff like Thingworks. And it is a rich, rich object modeling system with ton of configurable components on top of it. It's a GUI based system. teaches you how to build classes that then interact with each other, teaches you object oriented programming, but via GUI with dropdown menus and fields. just is hard. It's really hard. So this is where I think Fuse and others have gotten, know, ThingWorks started to 2013 and they chose not to use UA. They used an internal proprietary model, but the same goal, right? Like let's make models that

Glenn Gardner

Mm. Mm-hmm.

Sam - Litmus

Let's make classes that that self-described for data collection and system understanding and digital twins, but let's also have them be functional ⁓ objects from a software standpoint. can do things, I can ask them to do things and I can make custom work and I can call that custom work to occur. So ⁓ when Sesme got started in the early 2010s, ⁓ you could see on the market already the folks You either have two paths, right? You go with a standard based modeling convention. What else is there aside from UA for industrial space? Or you go with a totally proprietary. And ⁓ very much to Sesame's credit, they said, no, we're not going to reinvent the wheel. There's been enough of those already. ⁓ OPC UA has its complexities, but it is thoroughly thought through.

Glenn Gardner

Mm-hmm.

Sam - Litmus

They did this hard slog of sitting in rooms for years, figuring out like what worked well with DA, what didn't work with OPC complex data. In 2025, if we want to model a vending machine, how do we do that? They thought about all that stuff. And so the Sesame team really did a good thing and not take the opportunity to deliver to the world yet another standards-based modeling system. So the...

Glenn Gardner

Mm-hmm.

Sam - Litmus

complexities of UA really are like the working groups are slow. That's really it. And you can join the working groups and you can have all the influence you want, but it might be years before you see like a rest API get added with JSON and JSON LD instead of XML and this. And Randy Armstrong, UA OPC foundation, one man army of awesomeness. There's a lot to UA's wire protocols that make them great, but

Michael Finocchiaro

Okay. Okay.

Sam - Litmus

they're really harder to implement than what typically like HTTP interfaces that everybody finds ubiquitous these days require. So, you know, it's different as a developer, you can say like, I'm gonna go build this ⁓ interface to critical manufacturing data, subscriptions, history, all of that is a simple self-describing REST API with i3X. To get to that point with UA, with HTTP, you still have to do stuff ahead of that. have to, you know, you have different libraries you need to bring in because XML and then you get to these little things like JavaScript doesn't natively support XML. So how is the average developer working in JavaScript really going to read a XML? you like load in the side library community or you like string search, right? It's gross. So like it just.

Michael Finocchiaro

⁓

Glenn Gardner

Mm. you.

Craig

Mm-hmm.

Sam - Litmus

this doesn't have the barriers. didn't reinvent the wheel with another object model. And it's a really precise usage of the things that OPC UA got really. One of the things that UA does really well that is hard every, it's hard for everybody, right, Craig? Like class systems are really hard. ⁓ But it's a precise usage of just like the easiest aspects of HTTP integration. So yeah.

Michael Finocchiaro

Well, maybe we could jump back to that when we talk about day four when that presentation hit. Because I wanted to hear Dylan talk a little bit about that awesome whiteboarding experience that ⁓ Brad just mentioned, because ⁓ that was a really moving and interesting ⁓ piece of theater and education that you guys did. The whole red and blue was super cool. I'm glad that you re-explained it to me. The first time I was like, what? And then I was, OK.

Glenn Gardner

Mm-hmm. Thank you.

Craig

Okay. Thank

Michael Finocchiaro

Can you just talk a minute about that? I thought that was really, and it was like right in the middle of the show. That's what I really liked about it too. You know, it had two days of amazing content. It was going to two-millimeter, but you guys kind of framed right there. It was really cool.

Dylan DuFresne

Yeah, that one. mean, there's not too much of a story to tell there. I showed up at 730, got a mic on and got on stage at eight and hope people ask questions. I know on the conference side, they had to the whiteboard up there. I'm sure the AV team had a bunch to do, but I mean, as far as prep goes for Walker or myself, was live Q and A as it gets. was.

Glenn Gardner

Mm-hmm.

Craig

Thank

Michael Finocchiaro

You

Glenn Gardner

Thank

Sam - Litmus

I missed it Dylan, what were you whiteboarding?

Glenn Gardner

you.

Dylan DuFresne

⁓ This one is a pretty traditional ConnectCollectStore type piece there based on the questions that the audience came up with. So lot of the questions were around red and blue namespaces. It was how do we get started? A lot of the foundational pieces.

Michael Finocchiaro

But it. Well, maybe you could give us like a 30 second description of the red and blue, because maybe some of the people in the audience are like, what? ⁓

Craig

Thank

Sam - Litmus

Yeah, yeah.

Glenn Gardner

Yeah, that's what landed the best,

Sam - Litmus

Do you mean red and blue team? Like I'm gonna break your stuff, I'm gonna build it? Do you mean like accept messages, reject messages? Yeah.

Michael Finocchiaro

Or like American politics red and blue. ⁓

Craig

Condense that in 30 seconds.

Glenn Gardner

Yeah,

Dylan DuFresne

No. Yeah, so.

Glenn Gardner

know they didn't plan that. anyway, Dylan, in the audience, I was of course like cheering you on the whole way. the red and blue thing, the red and blue thing really hit well. I'll pass the mic back to you.

Dylan DuFresne

Yeah.

Michael Finocchiaro

Hahaha.

Craig

Thank

Dylan DuFresne

Yeah. a 4.0 solution coined term in this context as far as I know. Walker came up with this meaning of it and it's tangent tangent it's next it's Yeah, so it's basically to the left of your typical medallion architecture for any of the data science folks or people of that nature, your bronze, silver, gold, and the hierarchy of how you get that data out to the data science teams and the pieces of that. To the left of that is all the factor data at the edge that people like to ignore. And in the world of the 4.0 space, the OT networks and all of that, they call that two different types of data in there, red and blue. Blue data is,

Michael Finocchiaro

Okay.

Dylan DuFresne

I am an engineer, I'm in ignition or litmus or fuse, and I'm just making a random data model or a tag or a structure for my own purposes as an individual or a small team. It's not sanctioned by the company, it's not official, it's just I've built these data structures

Michael Finocchiaro

you

Dylan DuFresne

to solve my own problem. Maybe it's making a screen in ignition, maybe it's a data model to type pipe things into a historian or a database in a way that I can understand, but it's not sanctioned, it's not official, but it let me solve my problems. The red namespaces are the other side of that where it is the official namespaces. So the common example that you'll get from the 4.0 community is OEE and what is the data model for OEE. Every site, every area, every cell is going to have the same data structure for OEE or any other KPI, ⁓ probably some much more important KPIs, but OEE is the common example. So availability, quality, and performance, and these are the metrics. Here's how it's calculated.

Glenn Gardner

So

Dylan DuFresne

and having a structured enforceable structure for that model is red. The common misconception we talked a little bit about on the stage there that should be on the recording when that's live is a common misconception is that blue turns into red. But from a data science perspective, you can't have loose unstructured, ungoverned data turn into governed structured data. It just doesn't work. Once you lose that traceability and that quality of that data, things just become

Michael Finocchiaro

you

Dylan DuFresne

problematic the further away from that data you get. And so red and blue are kind of left and right, but they don't turn into each other. And then next to that, you've got your traditional medallion structure and much more traditional IT and data science terminologies for how we handle data.

Sam - Litmus

Yeah, Dylan, I wanted to add to that too. ⁓ a lot of this that where this comes from, our software development paradigms that ⁓ started in the late 90s in 2000. So when you have a big monolithic database

Michael Finocchiaro

I'm sorry.

Sam - Litmus

and the schemas are stored in the database, you're able to interact teams can push things back and forth into this big monolithic store because they know the table schema is the database is self-representing what the systems are. So when you have teams that are say like iterating on Amazon service, it has like nothing to do with like maybe Amazon, know, pet goods wants to ⁓ have a little bit of the data

Glenn Gardner

So,

Sam - Litmus

from consumer activity from Amazon video. They used to have to go to this, this internal database and check all these schemas and like, oops, didn't realize a schema change. I forgot to go grab it and then boo boo. So when Amazon started to

Glenn Gardner

Okay.

Sam - Litmus

realized that 400 people is a different company than 4,000 people, that database became an enormous burden for them. Huge dependency trees, enormous single point of failure. They weren't the first to do this, but they decided

Michael Finocchiaro

So,

Glenn Gardner

.

Sam - Litmus

to break out the teams, separate the teams, and then begin building from scratch infrastructure to facilitate it. This is a lot of the invent buses, the ideas that are now Kafka. This is a lot of the stuff where it came from. And when you do that, when you have one development team on one side of the house trying to build their stuff without dependencies on a single point of failure, but you want the other teams to be able to leverage each other's data, you have to have something to publish, you have to have something to subscribe, and you have to have something to display what the information is, what services are available from that development team's stuff. Like, hey, here's my REST API for my business, this is my data product, this what I do.

Michael Finocchiaro

So, thank

Sam - Litmus

I have services you can call to get my information from my dev team. can get the raw metrics if you want. You can get the code base if you want. Here's the information that this team is going to make available. So when you have modifications to those schemas and parts of these development teams, like really it's a distributed software application and a distributed set of development teams. The problem is the same though. How do you know each other's stuff? And so when you have a schema change, you have to handle the schema changes.

Michael Finocchiaro

So, you. you

Sam - Litmus

Changes to global schemas break everything. So you don't have a global schema. If you end up with what? These are red and blue, reject or accept. And so as you work in distributed software ecosystems, like in industrial automation, OPC UA solved that problem by having schemas stored in a repository that you go grab. So you could have these complex class type-defined relationships. And then just like Amazon, they're like, all right, yeah, you got to understand what this little entity wants you to.

Michael Finocchiaro

So,

Sam - Litmus

do with it. So go over here to this repository, right? But when somebody wants to make a new class or something, you got to validate it. Or if, oops, you start using it and pushing it onto the event bus, Amazon's like, no, no, no, that's useless info. We're going to reject the schema before someone approves it. So the idea is that we're trying to use inside of industrial automation right now, really solving two things. It's one, we need classes to better describe systems so we can call functions against them. We also need models and oftentimes classes are a great way to model. We need to represent self-describe the data too. And when you end up doing that and you want to have distributed systems, you have to have reject or accept schemas. It's a foundational principle of the software paradigms that now we are taking advantage of in industrial automation. The thing about UNS is that it's really two purposes. It's the shared backplane. It's the IPC, just like Amazon's IPC and it's a spot.

Craig

you

Glenn Gardner

So,

Sam - Litmus

for industrial data discovery. So we have a hard job. Like we have to facilitate the same thing that Amazon's doing. But we also like industrial automation plugged into a shared back plane, just like they invented EventBridge to have these distributed software development teams be able to share information and discover each other's stuff. We also

Glenn Gardner

thank you.

Sam - Litmus

have to use in the same way to drive equipment and drive machinery. And we need to make it AI ready. Like that's a really hard thing. But I think it's important that we just, we don't forget, we're built on the backs of giants. This is paradigm set by big data tech back when I didn't have gray hairs. there was a lot of, ⁓ all of Amazon web services, all those tools, are the represented, that's the tools Amazon built to build their forest strategy. And it's foundational to everything. It's cool.

Glenn Gardner

Mm-hmm. Thank

Michael Finocchiaro

So Wednesday. Yep.

Dylan DuFresne

jump on that real quick, the one problem with the red and blue namespace piece is in a full architecture where we stand on the giants, it's either I either liken it to be bronze the medallion or it's left of bronze. But in too many places where they just hear red verse blue for the first time, it was really a Stata and MES concept that was built.

Glenn Gardner

you

Michael Finocchiaro

you you

Dylan DuFresne

without full understanding of the data science side of things. And so too many people hear that and say, red is governed and then just make it UDTs in whatever skater platform they happen to be using and forget to do the governance piece. They forget about the functions. forget about.

Sam - Litmus

Yeah. I mean, mostly the governance that I know of are like the governance at the PLC structured data. Like that's it and then there's nothing above and you don't even get to sign colors to that. It's just like, did the integrator do it or not? Crap, they didn't do it. I got to go like piecemeal my register set again, you know?

Dylan DuFresne

right.

Glenn Gardner

Mm hmm.

Michael Finocchiaro

Okay.

Craig

Yeah.

Michael Finocchiaro

So ⁓ Wednesday night, you guys went off and did some medieval stuff. I was in the airport flying back to Paris. And then Wednesday, Craig, you were going to talk a little bit about Thursday, sorry, day four.

Craig

Day four. Yeah, day four was a little bit of a fuzz for us. We were up very late the night before redoing our entire presentation. And yeah, we went on, CESME was up first, right? And Sam and many others were a part of that kind of parade of folks as we went through the whole CESME presentation, learning more about the L5 3X.

Michael Finocchiaro

you But there was some Mr. Rogers neighborhoods. ⁓

Craig

⁓ It was something like that. I don't know

Sam - Litmus

Yeah, Matthew Ferris. was awesome. That was

Glenn Gardner

That was clever.

Sam - Litmus

awesome.

Craig

how

Glenn Gardner

That was.

Craig

you practice all those handoffs, but yeah.

Sam - Litmus

God. Right? That was really, I mean, he was like a, he was like a theater guy, I think. So he did some, yeah.

Michael Finocchiaro

That's it.

Glenn Gardner

That was legit clever. I liked it. Well executed.

Michael Finocchiaro

You

Craig

It ⁓ was choreographed well. ⁓ yeah, that was a huge drop, I think, for the show ⁓ and for the industry in general, right? To kind of get a sneak peek of what that's all about. ⁓ Obviously, as a sponsor and vendor, having that dropped on us at the last minute was ⁓ added an interesting twist as well. ⁓

Michael Finocchiaro

Okay. Yeah.

Craig

But yeah, like Sam said before, for those of us built on more modern tech, adapting that, know, GraphQL, ⁓ paradigms, quite straightforward and simple. So, ⁓ looking forward to working with Sesme more on those things. We had a lot of great discussions with John, you know, after that and after he saw our presentation. So we'll be joining you guys, ⁓ Sam, here pretty soon on some of those conversations. ⁓

Michael Finocchiaro

Okay.

Sam - Litmus

That's great.

Craig

Yeah, then we got

Sam - Litmus

That's great.

Craig

to go up on stage. We got to do our thing. So that was exciting. Like I've told a bunch of people, I've been doing this with Fuse for just about 10 years. This is the very first trade show that we've ever done as an organization. ⁓ Yeah. I've never... ⁓

Michael Finocchiaro

Wow. Drop out.

Sam - Litmus

Wow.

Glenn Gardner

I was shocked when you said that. ⁓

Sam - Litmus

Wow, no way. Congrats. Congrats. That was a special one. I mean, you've been to lots of shows though, Craig, right?

Craig

We've been, we've attended, and the problem always has been, I've just never really seen the value. ⁓ And we attended Prove It last year, 2025, and myself and Lance and a few others, and we all just immediately knew this is something we needed to be a part of. So it was kind of special for us in a number of ways. Just getting our engineers out of the office for one and bringing them somewhere where there was a lot of people You know, I think they were all nervous, but it went really well and we put a

Glenn Gardner

Can you hear anything?

Craig

lot of work in ⁓ So we enjoyed it we had fun up there very you know back-and-forth dialogue and we were really Happy that Walker came up at the end and kind of gave us an additional little plug. So that was very nice ⁓

Sam - Litmus

great.

Craig

I think we approached it, as Dylan said, from a very different perspective. We came top down. We really focused on the personas and the problems that needed to be solved. ⁓ Connecting and collecting, sure, but we kind of took it from the CFO's perspective. And

Michael Finocchiaro

Mm.

Craig

a lot of times in these big transformative projects, the CFO is the one signing the checks, approving the POs for these projects. We wanted to make sure that we

Michael Finocchiaro

Mm.

Craig

address those folks needs here as well. Because I think it goes beyond the control engineers desk and in a lot of this. ⁓ So we did that and, you know, from the plant managers perspective and all the way down, you know, how can UNS and having access to real time data through the organization really change

Glenn Gardner

There. Thank

Craig

the life of the folks that are living operations every day ⁓ outside of the control engineers office. So it was fun to be able to bring that to stage. and then right after us was the, fireside chat, you know, sponsored by litmus. ⁓ so that was a great conversation with some, ⁓ industry Titans there, you know, kind of going at it. ⁓

Michael Finocchiaro

Mm.

Craig

every, I had to go back to my booth for that. So unfortunately I missed the whole thing cause my entire team wanted to sit and watch that. So I was the sole survivor in the booth, but, ⁓

Michael Finocchiaro

Yeah. ⁓

Glenn Gardner

Okay.

Craig

Yeah, as the day went on, the other one that I attended personally was of course the Able Era session where I received a history lesson ⁓ from Glenn. It was amazing. ⁓ It was just the delivery. ⁓ I really appreciated what you guys did. You made it fun. You made it interactive. The delivery was informative in a number of ways. ⁓

Glenn Gardner

Okay.

Craig

you know, just the historical element of like how we got to where we are. I thought was really cool. And then I liked the way you guys presented the problem and how you work with companies to try to help them through those problems by giving them a solid roadmap. And unlike most consulting firms that want to kind of stay engaged forever, you know, giving them that roadmap and letting the organization take ownership of that and accountability for that I think is is a real big difference in how you guys approach that. So ⁓ yeah, that was all I really got to do personally that day. Those were my two highlights.

Michael Finocchiaro

Good highlights. ⁓ No, go ahead.

Glenn Gardner

Yeah, your stuff landed really, Fino, please go ahead. So I was going to comment on Craig's presentation. Yeah, like you're from the audience in your presentation, Craig, it landed really well. Like you guys had a blend of personalities on stage. So you and Lance are deep tech. Like it was clear to everyone in the audience that you and and Lance built this thing. You guys built the platform yourselves. The CEO himself, you is up there having a technical exchange informally live in front of the entire audience. Like that was awesome. It gave people the sense of what Fuse is. But then you had to got, you had your partners at PWC, you had Ed Sosnowski. And so like Ed talks like me, Ed is very familiar with organizational workflows and he can like stitch all the different pieces together.

Michael Finocchiaro

Okay.

Glenn Gardner

You and Lance can go as deep as Sam and Dylan can on deep tech. the

Michael Finocchiaro

you

Glenn Gardner

four of you are all kindred spirits. And then you have PWC to do the CFO and the CEO's office. was, it was a good combo, man. I, I thought it gelled well and it was totally informal. It was as though we were just like sitting in a dorm room or something, listening to four buddies talk. And I thought it was pretty cool.

Michael Finocchiaro

That's

Dylan DuFresne

And we can't forget that Fuse was the only one to bring a robot to the show. Everyone likes robots.

Michael Finocchiaro

true. ⁓

Craig

That's right. Yeah. Yeah. Yeah. So our, yeah, our partners, ⁓ yeah, FANUC ⁓ brought

Sam - Litmus

Yeah, yeah, I noticed that. That was awesome. That was awesome. And not cheap either.

Michael Finocchiaro

Oh Okay.

Craig

that out. You know, they, ⁓ they wanted, they want to be part of the conversation when it comes to industry four. ⁓ Oftentimes they get left out, even though, I mean, they're the largest ⁓ industrial robotics company in the world, but they get, they get left out of that because, you know, historically they've been focused on this legacy

Glenn Gardner

you

Craig

technology of everything is Modbus and it just works and that's just how they've kind of operate. We've been working with them over the past six months to help them level up that message, ⁓ getting them more into the real time, you know, message ⁓ facilitation between the devices. And so we've done a lot of work with them and we'll be going with them to IMTS. They've got a 4,000 square foot booth at IMTS and we'll be a part of that.

Michael Finocchiaro

⁓

Craig

connecting with their zero downtime solution. so, yeah, it's just, it's a different world for

Sam - Litmus

That's a great move, Craig. ⁓

Craig

us. Sam, you know, we're not the device driver guys. That's not our, you know, but as you get away from those proprietary drivers and you get into this open source community of MQTT and even OPC UA, right? It just makes everything more accessible to ⁓ all the companies out there. So yeah, we're excited.

Sam - Litmus

I mean, no, no, yeah.

Michael Finocchiaro

We only have about five more minutes. I'm thinking maybe we go around the table to see what we'd like to see next year, like what's gonna come. Like I hope that I'll be, I would imagine having a PLM platform there and being able to pull work instructions or the build process, the routing information. and make that part of the entire demo, for example. That would be kind of cool. Brad, what do you think? What would you like to see next year at Provence?

Glenn Gardner

Yeah.

Brad Hafer

Well, it's funny you mentioned that because we both share some of those PLM roots and we sat at lunch the first day and I sort of tasked you with like trying to figure out what are those use cases between the PLM world and this sort of OT world. And it's unclear. you know, not a lot is jumping out. That's this huge problem. So I think there's still things within this water column to use that term again for the proof of conference that needs to be proven going forward. One of the things that I observed overall

Michael Finocchiaro

Yeah.

Brad Hafer

You know, there's some big gorillas that were there, some big names, you very large, capitalized public companies, but there's also a lot of smaller companies. It's TBD, you know, TBD on who's going to make it and who's not. And the go-to market engines are still building. know, you know, Sam, like you guys have been well funded, but you know, Craig is still bootstrapping. And so there's a, there's a mix of vendors there and who's going to make it and all being capitalists, like you said.

Glenn Gardner

Okay.

Michael Finocchiaro

Exactly.

Sam - Litmus

Yeah. Yeah.

Michael Finocchiaro

Okay.

Brad Hafer

What's the ones, which horses do we want to bet on going forward? Because I think part of the challenge, not just in the 51 vendors that were there in the stack with the 1300 logos and the industrial software overall, it's confusing to customers. Which ones do they need to buy? Do they need one in every box?

Craig

Mm-hmm.

Michael Finocchiaro

Mm.

Brad Hafer

How do they connect them? Where do they start? Those are all challenges that the customers that were walking around the floor and sitting in the sessions are trying to figure out.

Sam - Litmus

Yep.

Michael Finocchiaro

I just hope the answer is not going to be just Zeeman's. I like the fact that they got more options. How about you, Dylan? What do you see coming in 27?

Dylan DuFresne

might be a bit more of a hot take, but I hope they pull back a little bit and stabilize. ⁓ Take what they learn in the first two years, build a foundation that can truly grow and build a conference that can last decades beyond any single personal brand.

Craig

you Mm-hmm.

Sam - Litmus

Yep.

Michael Finocchiaro

Nice. How about you, Sam?

Sam - Litmus

Yeah, first off, this has been a great podcast. Thanks for having me. I'd come back and maybe you guys want to come on ours. ⁓ But so thank you. But ⁓ for next year, I'm just going to go full on industrial automation geek. ⁓ Just, you know, self-describing automation controllers, automated integration, ⁓ end to end, ⁓ you know.

Michael Finocchiaro

You're welcome. F. Hahaha!

Sam - Litmus

let's see the PLCs explain to the rest of the business how they want to be integrated. Let's see the business systems in the middle ⁓ know what to do with them. I think from a tactical standpoint, I'm excited about showing more of that.

Michael Finocchiaro

How about you, Glen?

Glenn Gardner

There's three communities that I love that I'm a part of, and this represents one of them. This is like the OT community. Fino, you're a leader in the PLM and engineering design community. It's right next door. Those two teams really need to talk to one another and my roots.

Michael Finocchiaro

Instead of this tossing the bomb over the wall, like we've been doing.

Sam - Litmus

Like a digital thread needs to be created maybe. I'll put on my PTZ hat real quick. Hold on. Sorry.

Glenn Gardner

There you go, there you go. and predict it. So today there's, that's all good, there's three awesome communities. One is in the predictive maintenance world. I'm sorry, say it again, Fino.

Michael Finocchiaro

What's the third one? PN. Okay. No, you said there were three. There was the OT, there's PLM and then number three. good maintenance. Maintain X.

Glenn Gardner

And then predictive maintenance, call it maintenance technology, predictive maintenance, CMMS. And yeah, so you see the inklings of them coming together. So maintain X is a cornerstone of the maintenance and reliability community. Love those guys. ⁓ but they're the only ones that are showing up at this OT stack. Fino, you are, don't get a big head, but like you're a cornerstone of the PLM and the CAD community. And you came to the show too, which was awesome.

Michael Finocchiaro

lesson. Thank

Glenn Gardner

I'm really wrestling with, should we bring these three communities together? Should they continue to exist separately? I know for sure I want to mix them more because when you're on the plant floor, all three come crashing into the same situations. So that's what I'm most excited about. PLM, OT, maintenance technology. I want more mixing of those three.

Michael Finocchiaro

it. Awesome. Just a little side note. I just posted a thing about maintain X and the interview that I did with Alex. So go ahead, Craig, how about you for next year?

Craig

⁓ Yeah, I second everything everybody's already said. You know, I think I've got a couple ideas I want to share with Walker, just a couple different concepts. I liked the workshops. We didn't really talk about the workshops much here on this podcast, but I thought those were really valuable for folks to be able to see a little bit more under the hood because the presentations, they're so polished. They're so practiced, right? They're so rehearsed. ⁓ And then you get into the workshop and it's just very raw. know, it's you get

Glenn Gardner

Yeah.

Craig

questions out of left field. So I really liked that. I would like to see a little more of that. I would actually like to see, you know, maybe Sam and I and Travis from Michael Finocchiaro (1:00:01) Thank you. Craig (1:00:07) Ignition in a room for half a day and we build a solution from scratch. And people can come and watch that happen in real time. I think that would be, yeah, like a build-a-thon or, you know, something I think that would really draw, I think a lot of people to, we had a lot of people come to the booth. They're like, Glenn Gardner (1:00:07) Nice. Michael Finocchiaro (1:00:16) kind of coding in midair kind of thing, right? Craig (1:00:27) This all looks amazing, but how hard is this really? You know, so I think that would be cool. That's just my recommendation for Walker if you're listening. Michael Finocchiaro (1:00:30) All right. Hahaha Glenn Gardner (1:00:37) I'd pay to watch that. To throw like you, Dylan, Sam, Travis, Lance, like, through those five guys on stage, I would watch anything you guys did. Michael Finocchiaro (1:00:39) He threw them. Craig (1:00:40) Yeah, we can sell tickets, right? Yeah. like, share our screens up on the big screen as we're doing it, you know, live, like, and just throw hurdles at us, you know? Let's see how it really performs. I know that's like crazy, but that's just the CEO talking that I want to solve those problems. Michael Finocchiaro (1:00:56) That'd be cool. Glenn Gardner (1:01:02) You Michael Finocchiaro (1:01:03) Dylan, do you have any clue in the other videos are going to be up there supposed to be up today or tomorrow next week? Dylan DuFresne (1:01:09) The AV crew, they've got many different, multiple terabytes of video footage to go through. Last I heard it was at least 30 days for the AV team to go through it and then things should start trickling in from there. Michael Finocchiaro (1:01:15) Okay. Craig (1:01:15) Yeah. Michael Finocchiaro (1:01:21) Okay. All right. Well, stay tuned for that everybody. ⁓ Thank you for ⁓ joining Dylan, Sam, Glenn, Brad and Craig. It's been absolutely fantastic. And it's been awesome for me and a privilege to join your world for a week, a little more than a week now and kind of understand the problems that are very similar and different than the problems in my side. ⁓ For those that are listening, Glenn Gardner (1:01:24) Thanks. Michael Finocchiaro (1:01:49) There'll be another podcast in about an hour from now with Cosmon and ⁓ Metafold 3D, two really awesome simulation Glenn Gardner (1:01:56) you Michael Finocchiaro (1:01:57) companies. ⁓ And once again, thank you to everybody. I'll probably have Craig on another podcast. I'm sure I'm going to have ⁓ Brad and Glenn and Sam and Dylan on as well. ⁓ But it's been an absolute pleasure today and thanks everybody for joining and we'll catch you next time. Craig (1:02:13) Yeah. Brad Hafer (1:02:14) Thank you. Craig (1:02:14) Yeah. Thanks. Glenn Gardner (1:02:15) Thanks, Sam - Litmus (1:02:16) See y'all. Bye.