Peter Zornio: Emerson's Boundless Automation Vision

Episode 92 • November 16, 2026 • 00:40:36
Peter Zornio: Emerson's Boundless Automation Vision
Ayna Insights
Peter Zornio: Emerson's Boundless Automation Vision

Nov 16 2026 | 00:40:36

/

Show Notes

In this episode of the Titanium Economy podcast, Peter Zornio, Chief Technology Officer at Emerson, talks to Vineet Gupta, President and Head of Semiconductor at Ayna, about how Emerson is building a full-stack automation platform for the modern industrial economy. He explains how the company's boundless automation vision connects sensors, control systems, edge computing, cloud technologies, and enterprise software to help manufacturers improve productivity, reliability, and operational performance. Peter also discusses how AI is transforming industrial automation, from preserving the knowledge of experienced operators to predicting equipment failures and enabling an "ultimate operations advisor." The conversation explores why a strong data foundation is essential for industrial AI, how Emerson approaches AI in safety-critical production environments, and how its technologies are supporting semiconductor fabs and AI data centers. Peter also reflects on his four-decade career in automation, the role of engineering and innovation, and the skills the next generation of engineers will need to succeed.

Peter leads innovation strategy, product roadmap, central R&D, and industry standards across Emerson, the roughly $18 billion St. Louis-based automation leader whose technology runs inside the plants that make fuel, medicine, chemicals, and power. He joined in 2006 and led development of the PlantWeb digital ecosystem before becoming company-wide CTO in 2022, following 21 years at Honeywell. He holds five patents, serves on the FieldComm Group board, and earned a chemical engineering degree from the University of New Hampshire.

Discussion Points

Ayna is a premier advisory and implementation firm in the industrial technology space, leveraging a team of experienced leaders to help companies and investors drive performance improvement and value creation. The host of this episode Vineet Gupta is President and Head of Semiconductor at Ayna.

For More Information

Peter Zornio LinkedIn

Emerson

The Titanium Economy

Ayna Website

View Full Transcript

Episode Transcript

[00:00:03] Speaker A: Welcome to AINA Insights where prominent leaders and influencers shaping the industrial and industrial technology sector discuss topics that are critical for executives, boards and investors. INA Insights is brought to you by INA AI, a firm focused on working with industrial companies to make them unrivaled segment of ONE leaders. To learn more about INA AI, please visit our website at www.aina.AI. [00:00:40] Speaker B: Welcome to another episode of Titanium Economy podcast hosted by ina. Our guest today is Peter SVP and Chief Technology Officer at Emerson. Emerson is a global automation leader with annual sales approaching 19 billion. Its technology sits inside plants that make our fuel, our medicine, our chemicals, our power, and increasingly inside the data centers behind AI. Vidu is one of the most respected technologists in the industry. He spent four decades in automation, 21 years at Honeywell, 20 years at Emerson. He joined in 2006 and led the development of Emerson's system and software portfolio including the PlantWeb digital ecosystem and became Emerson's company wide CTO in 2022. Today he leads innovation strategy, product roadmap, central R and D and industry standards across the entire business. He's also given a great deal back to the industry itself. He serves on the board of fieldcom Group which manages the key standards in industrial automation. He holds five patents and has a chemical engineering degree from University of New Hampshire. Peter will retire at the end of the year, which makes this a special conversation. We get to hear about what he has built and what he sees coming next. Peter, we're delighted to host you on the podcast. [00:01:53] Speaker C: Well it's absolutely great to be here. Thanks for coming. [00:01:56] Speaker B: Perfect. So Peter, Emerson was founded in 1890. It's been 135 plus years and the technology that Emerson has is used in plants that make products that we use on a day to day basis. But not a lot of people know about immersion. Can you tell us what does Emerson do? [00:02:13] Speaker C: Certainly. So Emerson is what we like to call a full stack automation provider. So in a special essentially the process in hybrid industries, the one where chemistry and flow play a bigger role like some of the ones you mentioned, like pharmacy or refining or petrochem, we have all of the technology from the sensors, the control systems that take that input in and figure out what to do with it and then manipulate the final control elements, valves, actuators to achieve the complete control loop as we call it, right. To make sure that those facilities are running in control safely and on target. We also have software that lives above that that optimizes, connects that with ERP business systems, figures out what should be going on in the facility to optimize the profitability or the reliability. We also now are in the test automation business since we acquired a company called National Instruments a while ago. So you know, really have kind of a unique position in automation by having that complete stack of technology from top to bottom in the discrete industries, which is more mechanical manufacturer. We also have a portfolio not quite as complete as we do in the process in hybrid industries. But you know, everybody who makes things, they're pretty familiar with us. But the consumers that use them, not so much because yeah, they're not, they're not there in the factories making the products they're using. [00:03:47] Speaker B: Yeah, Emerson has been through a remarkable last four to five years. You sold the climate and the consumer business, you acquired National Instruments that you just talked about and you bought aspentac fully in house. And today you are one of the most complete automation portfolio in the industry. From sensors in the field all the way up to the enterprise optimization software. From where you sit as a cto, what can Emerson do for a customer today that it couldn't do four to five years back? [00:04:13] Speaker C: Yes, exactly. As you mentioned, when Lal Karasambhai, our CEO took over, it was his strategy to divest non automation oriented business so that we could build out this full stack and basically do something that the industry hasn't seen before and have a unified suite of software and applications from top to bottom. You know, historically automation vendors have been good at one piece of that stack, whether it's the instrumentation or the control systems, the optimization software. But the opportunity exists to tie that all together in a unified way that will make it far easier for them to deploy support cybersecur. I look at it very much as the shift that if you're old enough to remember happened in the ERP world when it used to be individual software companies for accounts receivable or human resources. And now everybody's so used to that space being dominated by integrated suites. Well, we're at the position where you can actually achieve the same sort of thing, but for all of the automation and what we call OT operational technology functions that really make the manufacturing happening underneath those ERP systems, [00:05:25] Speaker B: that's perfect. And I think that's a segue. To my next question, let's talk about boundless automation. And I think you've carried that vision since 2022 and it has held up pretty well. I mean you looked at the architecture that the whole industry has used for 30 years and said that the data needs to move freely across the field, the edge and the cloud. Take us back to that moment. What were you seeing in a customer plan that told you that it was a time to rethink the whole thing? [00:05:50] Speaker C: Yeah, I think the first and foremost thing was the technology shifts that were happening. Right. So in the field there was advanced technology occurring in terms of higher speed digital interconnection for the field devices themselves. There is the next generation of wireless devices coming out that we're very engaged in, in the kind of middle layer, if you will, of the architecture with the control and automation systems. You know, edge technology as it's called, was becoming the Dominan sort of force there. For years that part has been based on Windows operating systems which have served us very well. But as we go forward we also see this ability for different operating systems, for using containerization and actually having a much more flexible compute environment on premise the same way that basically cloud compute environments. So those technologies from the cloud coming into the on premise world and then the cloud itself is, had gotten far more accepted. You know, the operational technology folks traditionally viewed cloud as something that IT people did and you know, we don't trust it. That's something, you know, we want to have control of our own systems. That attitude had changed quite a bit. So when we step back and we look at it, we said, you know, it's really going to be digital, smart digital devices at every level of the architecture, talking to each other. We recognized that we needed also what we call now a data fabric to tie all that together. So I think we were looking at those technology shifts primarily first. That drove us into realizing that this is an opportunity for a big change. The second thing that happened of course is we now found ourselves with all of those pieces. So the logical thing would be, well, let's put all of those pieces together in an architecture, right? So it was a good combination of I'd say the technology shifts that we're having in the industry at large with the portfolio expansion we had that said, well, this is kind of to me obviously the path we need to be on. [00:07:56] Speaker B: Yeah, no, that's perfect. And I think your vision is now a real product. You know, Project beyond has become an enterprise operation platform. Delta V16 shipped in January. AspenTech launched in May. Innovation AI agents begin deploying in this fall. So explain it the way you would to a plant manager running a 20 year old control system. What does he get and how does he protect the investments he's already made? [00:08:18] Speaker C: Sure. So for someone who's got older legacy digital automation, which is from the 80s and 90s now, which when I was a young engineer was the state of the art stuff. So I'm looking at stuff from the beginning of my career now as more legacy. We've always had the ability to preserve some of the physical assets, the hardware, the wiring, by coming up with special connection mechanisms to those older systems. What's really changed in the last few years is we've built a suite of AI tools that does what's the hardest part, and that is taking the specific plant level, level, configuration and translating that into a modern control system. Right. Because all of those systems were proprietary. They all had their own proprietary languages, configuration schemes or configuration tools that were used. So it used to take an engineer who knew both systems, the old and the new, and many hours of integration work to convert an older system to a newer system. This is no different than anybody who's tried to move a legacy system in banking or any other industry into a modern system. They know that's a hard task and a lot of work, and it's viewed as risky because you need to do that while you're continuing operating. But with these AI tools, we were able now to do over 80% of that work all automatically. So one of the big barriers that had always occurred, that cost of making that transition, we've been able to remove. [00:09:49] Speaker B: And I want to dig a little bit onto that particular piece as well. And I want to start with one of the lines. You said, right, that Emerson was doing AI before AI was cool school. And when I read about it like this, you have been using machine learning on control systems on a live refinery to make adjustments live. And you have had some of these products running since mid-1980s. And so now you actually have a four decades of perspective that nobody else has in this particular conversation. Right. Which is pretty interesting and very cool. So what is genuinely new today and what has quietly been working for the last 40 years? Can you share that perspective? [00:10:26] Speaker C: Yeah, absolutely. It's. It's something I think about a lot as we plan our future. So the AI we've used, you know, for now over three decades, is traditionally based around numerical methods, neural nets, machine learning. They were models that were built around very specific application areas. So every model was for, you know, this is a model for optimizing production. This is a model for analyzing reliability and equipment. Okay. That was the state of the art of AI at that time. But now as we look at large language models, the multimodal nature of that, the fact that you can bring in all types of different data into those other than just Numerical data. As we look at agentic AI and the ability to tie those together, there's a whole new world that's opened up. I would say the first set of things that we've done is even with those models that we have today, just being able to speak to them, you know, to figure out, to tell them what you want to do or ask them to give you a projection of what would happen has been a great new interface. We've used that technology also to make it a lot easier to use the technology we have through configuration tools and advisors. The big prize is really what we call the ultimate operations advisor. You know, this is the. This is something that Hollywood has already set the expectation for over many years. You know, I would say going all the way back to A Space Odyssey 2001 or, or more recently, if you look at the Iron man series with Jarvis, what our customers are thinking now of is, well, how do I get an online system that knows how my plant works, knows what's happened in my plant in the past, has the knowledge that a very experienced operator might have about the very specifics of that plant, and is able to provide at least guidance to today's operations team. And eventually the vision is always closing the loop and autonomously potentially doing the things the operations people were doing. One of the things that our customers hate to see is very experienced people retire because they know there's a tremendous amount of knowledge leaving that's in their heads, not about how to run the plant when everything's fine. We do a great job of that with today's automation, if things are at steady state. But what do I do when something abnormal happens, when the piece of equipment fails, if the weather turns into a situation that impacts the manufacturing, if raw material changes happen, or if you want to change grades or the product mix you're making? That's where those experienced people are really valuable. That's where we've struggled to automate making all of that happen. With today's AI, we have a real shot of doing that. It's much easier to do that today because we can take all of the data sources that that experienced operator would use, like the knowledge in his head about what happened in the past. A lot of that's in logs and in history engines, standard operating procedures for that plant. Okay, those are written down. You know what those are. Okay. What to do in terms of reliability when something fails. Okay. We have cmms, computer maintenance management systems that have a lot of the ass information in them. Today. It's been taking a human to go across those vastly different silos of information and basically synergize them into a view of what should happen. Today's AI models are very good at doing that. That's what everybody sees. And they're good at doing it without you having to describe every scenario and thing that might happen and, and then say, well, this is what we should do if this happens. You can just feed all that in and let those interactions between the different conditions that occur be inferred by the AI model. So we really see an opportunity for a step change in more than automation. What the operations advisor would be able to do by being able to go across all those different data sources and having a great view of what should be done at a very plant specific level. Because most of the time our customers, no two plants are the same. So that's another reason why it's been difficult to roll some of these things out in a broad way. [00:14:55] Speaker B: So actually the point that you talked about, right, it's actually a genuine concern at many industrial companies that I interact with, which is the old experienced operators are retiring and there is a digital native generation that is coming in. And with that whole, you know, guard retiring, there's a lot of knowledge that was not documented, but it's in their head. It's also walking out of the door. Right, you talked a little bit about how you are helping, but in that specific, in these specific cases, can you talk about like a case example where you have helped your customers retain that knowledge and so that, you know, there is no discontinuity in the operations? [00:15:36] Speaker C: You know, I have been involved in many different software programs to try to help exactly that. Whether it was from abnormal situation management where, you know, you try to say, okay, we're going to find what the root cause of an alarm flood is and make sure, you know, the operator knows how to deal with whatever the particular incident is. In the 90s I worked on, you know, what we called, you know, knowledge management systems back then. We even had specific languages that those were coded up in. We, we have software today that we call state based control that manages transitions. Those work, but they take a fair amount of effort, you know, to actually codify what needs to be done and under what conditions that action should be taken. So there are more deterministic, they're more cause and effect kind of structures. So they're not flexible when they see new stuff. The value here that we see is with AI engines today, being able to go back and look across potentially decades of past history and see those correlations and do that automatically without you having to codify that upfront. So that's. You know, we've released an advisor for our Ovation control system. We have one coming out next month for our Delta V control system. Our AspenTech folks have put that together around their applications. That ability to do that automatically without a whole bunch of upfront configuration work is what the real promise of today's AI is. [00:17:16] Speaker B: Yeah, And I see the value in this one primarily. Right. Because uptime is critical, especially in the process industry. When the line is running continuously, it is even more critical. And being able to get to the right answer in a much shorter time, which I think some of your solutions are enabling, is the real dollars and dollars, both in revenue as well as in margin for the customers that you're serving. [00:17:38] Speaker C: That is exactly correct. And another big area that we expect to see big gains in is reliability. So some of the industries we serve, that's their single biggest profit opportunity, their biggest profit loss is unplanned downtime. So if we can get rid of the unplanned downtimes through ever better predictive equipment, diagnostics and reliability, which we've been working for a long time with machine learning, but now we have the ability to pull in other data sources to help, you know, and again, automatically have the engine make that correlation between all those things, that can be a big help also for a lot of our customers. And today our customers are faced with not just how do I maximize production for optimum profit, whether that's lower cost, more production, or maybe flexibly changing to the optimum set of products. Okay. That's been a traditional goal. Reliability has always been a goal. Now we have sustainability as well. So you have all of these operational goals, and really, you need to optimize across all of them. They can potentially be in conflict with each other. Okay. So the scope of what you need to optimize across has also gotten a lot bigger. And that's again, an area where the more complete models that we can put together and the ability to actually see how those things interact is going to be a big help. [00:19:01] Speaker B: Peter, one of the other things you said was there is no AI strategy in manufacturing without a data strategy. And I see that. Right. A lot of AI programs get stuck in the pilot stages, but many of your customers are past that stage. So when you're having the first conversation with the leadership team of your new customers, what do you tell them? [00:19:18] Speaker C: Yeah, we tell them that. Okay. And we often start, you know, the two become combined. Oftentimes, the AI Strategy and the data strategy. Right. So, you know, we just recently had a very, you know, successful engagement with Total, the large energy, French energy company, where they also had the same vision around. Okay, we have to have a data strategy in place for whatever AI things we're going to do in the future, some of which we have programs that we know of today, but ones we haven't even thought of yet. Right. So a lot of times, you know, it takes some convincing with the customer to realize like, okay, that that data strategy is going to be a key underpinning. You know, you might be really excited about the sports car, but if there's no roads, okay, for the sports car to drive on, it's not going to do any to go straight to the sports car. So we often have those discussions at the same time. And whenever I talk about this topic, we just did a number of customer events. I always start with like, okay, Data foundation. And then that enables all this AI benefit that we're talking about, such as the Ultimate Operations Advisor. [00:20:31] Speaker B: Perfect. Now actually switching topic a little bit to when you actually go and implement some of these technologies. Right. And I want to talk a little bit, I want you to talk a little bit about Emerson's engineering discipline. Because when you implement some of these solutions, the AI solutions in a live production environment, safety uptime are very critical and you cannot sort of compromise on that. You cannot put a solution and patch it later. Right. Like it has to be right from day one. Can you talk about, you know, how that engineering discipline and how do you approach putting the right solution the very first time [00:21:06] Speaker C: we kind of sum that problem up as hallucination is not an option. Okay. Our customers are very used to deterministic solutions. Even the machine learning driven kind of solutions we've had in the past are deterministic. So there's a number of things that come into play. First, pretty much all of the operationally tied solutions we're looking at run on local LLMs. The customers don't want connections to the Internet for cybersecurity reasons and also because they still have some fear and paranoia about their domain specific knowledge of their facility leaking out. Okay. The second, and this is probably the most key thing I would say, is that what goes into those models has to be very carefully scrubbed. You know, it has to be very carefully curated to make sure it's accurate. So if you're putting in a standard operating procedure for a particular facility, it better be up to date in the correct one. Okay. Not something that was written eight years ago and you know, changes have been made in the process since then. If you're putting in, you know, historical data, you know, you better make sure that, yes, that's data that I trust to be accurate. You know, so curating the data that goes in is extremely important because it's that data that we're going to build, the models around that particular facility is operating. We don't want whatever frankly knowledge might be in the LLM already about the process. That's not why it's there. Okay. Because we're going to build a much better, very specific knowledge to that facility than some generic knowledge an LLM might have about. Here's how you start a boiler. No, no, it needs to be the very specific process for that facility. So the data that goes in has to be well curated. Okay, but even with all of that, you still expect, customers are going to expect some explainability, which is another term you hear a lot in the AI world. And it's true. Wherever we can, you got to show them why it's telling you what it thinks it's trying to do. Now the more sophisticated and complex the model gets, the harder that is to do. But when we do have direct correlation like here's the next step in the startup, we can point to the actual sop, but the last thing I would say is we can do all that stuff. It's going to take experience with these systems before customers trust them. That's just the way it is. Even with apc Advanced Process Control, Model based Advanced Process Control, when that was first going in the, you know, mid to late 80s and 90s, operators would be like, ah, what is that crazy thing doing? It's trying to like, you know, move the set point over to here. I don't believe that I'm going to turn it off. And so, you know, we've, we've seen this problem before of it's got to as much as possible, the operating team has to believe in the advice that they're giving and that's going to take a few years of living with the software before it really fully has their trust. [00:24:20] Speaker B: Perfect. A lot of our listeners come from the semiconductor world and there's an enormous wave of, you know, semiconductor fab construction going on. Of course, the semiconductor fab is one of the most demanding sort of environmental conditions, right? Both in terms of the atmospheric conditions as well as the ultra pure water specialty gas. And this is an environment where even a couple of or a fraction of a percentage point changes in yield would cost millions of dollars. So where does immersion show up in a modern fab. And what does the NI test and measurement business add to that story? [00:24:53] Speaker C: Sure. So we show up really in two very different areas. First of all, in a modern fab, if you go to the basement of that modern fab, it looks like a chemical plant, right. As you just described, the water processing, the gases that are used, all the stuff that goes in there and there, all of our traditional automation equipment applies, our sensors, our control systems, whatever. So you go up into the fab process itself. Well, that's where test and measurement, as we call it now, the former national instruments plays a big role. So they play a big role way back in the process of actual validating the design and the process for making the design, which is kind of their traditional heritage in the semiconductor world. But now they're also very much more increasingly involved in production test. So having units that as the wafers are built or even as they get packaged to come out, doing the validation on the quality and the test on that, for instance, they have a piece of software called Optimal plus that can look and analyze an entire wafer for quality and provide feedback into how the manufacturer of that entire wafer could be approved. Using analytics, machine learning, kind of analytics to trace where are we consistently seeing the potential defects in that way for back to where that might be happening in the manufacturing process. So yeah, we're definitely in that space. That space is having a great run up this year. Right. It's kind of a classic sort of boom or bust industry and we're happy that it's now kind of back in the boom phase. We just did our earnings release this week and you know, you can go look those numbers up, their public numbers, and you'll see that that space is, is way up for us this quarter. That semiconductor space, as reflected by the numbers that we're reflecting for our test and measurement business. [00:26:44] Speaker B: Yeah, Another sort of parallel space is the AI data center. And the build out needs exactly what Emerson is good at. Right. There's a precision control of power and cooling, the industrial grade reliability, and there's getting enormously complex project build fast. How did that opportunity find you and how big do you think it will become? [00:27:01] Speaker C: Yeah, so again, two very kind of separate but related areas that we play in. So the first way really that it found us was as people are building data centers. Power is of course a big question. The way a lot of them are solving it in the US at least is by generating their own power. Right. So that behind the meter power, as it's called, well we are the leader in power generation optimization. Over half the power generated in North America is controlled by our systems. So it was a natural for us to be the company they all turned to as they were doing their own power generation with gas turbines or solar or whatever they might have been doing as part of actually building the data center. From that, we also were able then to say, hey, inside the data center itself, you've got a lot of cooling water that needs measurement, that needs control. With control valves, isolation valves, you need some systems like SCADA systems that monitor all of the data center racks and give you a view of what's going on. So we certainly are very present in the power generation side of what's going on there. And we're increasingly seeing more and more business in the actual data center itself from the cooling water and utilities perspective, as well as monitoring the data center. [00:28:25] Speaker B: So let's talk about this AI revolution a bit, right? Immersion sits at the intersection of industrial economy and AI revolution. And if everything that you and your team have forecasted to do, and if that pans out, what does American manufacturing looks like in 10 years? [00:28:41] Speaker C: Yeah, the biggest payoff is very simply expressed in one word, productivity. So if you look at American manufacturing, right, we have to compete largely on productivity. We don't have the lowest labor costs, not always the lowest raw material costs, although we do when it comes to natural gas products now, which has been a great boon for the US And Canada. But in general, when automation, both for discrete, hybrid and process industries, when digital automation first hit the scene, first started mid-70s, really strongly came on in the 80s and 90s, American manufacturing, global manufacturing that took advantage of it, saw a big productivity hike. I mean, I remember talking actually with a multinational refining company, you know, probably around 2000, and they were looking back and showing us, you know, hey, across our refining fleet, with applying all this digital automation, we got enough extra capacity that we didn't have to build another refinery. And you. And that's proven out by just the big statistics. If you look at manufacturing productivity, and I think that first wave of digital automation had a huge role to play in that. If all the things I just talked about, if we're able to capture the knowledge of how to deal with any kind of abnormal situation happens in a plant, if we're able to predict reliability failures before they occur and dramatically increase the uptime of the facility, if we're able to much more flexibly change to market demands or whatever, again, driven by these AI solutions with our optimization and digital twin software as underpinnings That'll be the next big kick. I think that is the next wave of digital technology that will provide the same kind of kick we saw with the first wave that happened in the 80s and 90s. So the big prize is very simply productivity. [00:30:55] Speaker B: No. Interesting. And I couldn't agree more with that. And that actually be a big boost. Peter, I want to shift gears a little bit. I want to talk a little bit about you. [00:31:03] Speaker C: Okay. [00:31:04] Speaker B: So you started as a chemical engineer and trained at New Hampshire and then you went straight into automation. [00:31:09] Speaker C: Yes. [00:31:10] Speaker B: And you've been there ever since. You know, first close to 20 years at Honeywell and then close to 20 years at Emerson. In this industry, especially in the industrial economy, people change careers every three to five years, years. But you really found something worth staying. What was that? [00:31:25] Speaker C: Yeah. So for me, I think I was very fortunate because I always knew I was or pretty much knew since I was young that I was going to be an engineer and then pretty much figured out I was going to be a chemical engineer as I played with my chemistry sets and made model rockets and did a lot of chemical stuff. But then when I got into college, that was the early days of computer science. And then I discovered computers and I'm like, wow, I really like this too. Okay. You know, this would be another good avenue to go down. Well, automation is really the combination of those two skill sets. Right. The chemical engineering background really helps me understand our customers processes, what they're trying to do, what their points are for increased profitability or production. And then the knowledge of computer science, networking, software, all of that is how does that get applied to make that happen. So I think I was fortunate to basically find a space, not a space that people study, at least in the United States directly. Right. We don't have undergraduate degrees in automation in the US you can go to a number of more trade or technical oriented schools that will do automation in India and other countries. They actually do have undergrads in that. But I was lucky to land in that space. That was probably a combination of the two most technology oriented areas that I was interested in, chemical engineering and computers. So once I got there, I stuck. And I was fortunate, I think, because I started as an application person, applying automation, doing some advanced control applications. That gave me a great background and understand the customers. And then from there I realized what I wanted to do was actually figure out what those products did. I still remember sometimes I'd be working on like, who designed this? Don't they know how we really use this? I want to be the guy that actually drives the direction that the company goes in terms of what products we make, where we invest, what the next thing's going to be. So it's worked out well for me for over 40 years. [00:33:34] Speaker B: So over the last 20 years, you've been leading Emerson's innovation engine. And building something that does not exist takes a very different kind of leadership. How do you decide which bets you want to take? And how do you motivate great engineers to go after a problem that has not been solved? [00:33:50] Speaker C: Yeah, that's a great question because it often just, quite frankly, also takes time. Okay. Motivating the engineers is definitely one aspect of it. But what you find is energized. Engineers who really understand a space don't need much motivation. They want to go work on the next thing. AI is a great example. I mean, we have some fantastic people inside our different business units, product lines that are fully engaged. I mean, they work on it in their spare time. Right. Anything that engineers are really interested in, they continue to bring home with them and work on in the weekend and teach their kids. I've leveraged that now by trying to put them together in a team to spread the enthusiasm, the evangelism, if you will, across the different product units. Almost tougher, I'll just be blunt, is keeping management on board with the level of investment to do these bigger, longer term things, you need to really be able to cast a good vision. It's an often overused, but still very true expression. Let's go to where the puck is going to be, not where the puck is, you know, as the great one, Wayne Gretzky, you know, said in hockey. And that's really what the game is in innovation. You have to be able to read the tea leaves of, you know, where the technology is going and what the potential is going to be for that particular technology to be integrated with your space to deliver new results. Much the same way I just talked about how we could use the new AI technology to deliver all kinds of new results for our customers. A lot of that is experience. A lot of that is working with customers for a long time to get a feel again, what their needs really are, how they're going to respond to a product when you have it out there. And fortunately, I've got quite a bit of that, as we just discussed. [00:35:41] Speaker B: Perfect. Peter, one last question. You retired at the end of this year. [00:35:45] Speaker C: Yes. [00:35:46] Speaker B: And you've spent four decades leading the automation work in this industry and setting standards that everyone lives by in this industry. What are some of the things that you are most proud of and also, what advice will you give a new engineer who's just starting now? [00:36:01] Speaker C: Yeah, I'm probably, I would say I'm most proud of two things. The first one is, I think, a very specific product introduction. New technology, new product introductions that I was a part of. I was never the core developer, but I was in many cases the chief cheerleader and evangelist for it amongst our customers after they came out. Frankly, both at Honeywell and Emerson, I'm still on Emerson's payroll. So I won't pick sides and talk about specific ones. So I'm very proud, as I think back of some of those products that I look at and I go like, wow, that was a fantastic introduction. Okay, I will pick one. When we did electronic marshaling and charms here at Emerson For DeltaV, that was really a great innovation that was very easy to tell customers about what the value was going to be. The second thing is, again, maybe a little trite, but the impact you have on other engineers. I mean, I am getting wonderful emails now from people I've worked with here at Emerson, like, hey, you really helped me here, you inspired me to do this. Or I'll never forget the conversation we had about that. And that's just priceless. Right? That just makes you feel really good. [00:37:15] Speaker B: Yeah. So, Peter, what advice would you give to a new engineer who is just starting out right now? [00:37:20] Speaker C: Yeah, I should answer that part of the question as well. So I think that it's pretty clear as we're looking at things, it used to be a very solid career would be like, okay, go into computer science, learn to be a good programmer. Well, you're going to have to be a lot more than that going forward because a lot of the core programming is definitely being taken over by AI. You have to be very good at software architecture. The other bigger thing I would say is you have to understand a domain. You have to really be able to understand how things are being used, what the market requirements are going to be. As AI is going to more and more facilitate the actual development part of products moving towards, you know, what the industry calls spec driven development, the valuable skills are going to be increasingly being able to tell AI what to do or to review or to be able to have that feel around, like, yeah, this is the kind of product that will succeed in the marketplace. You know, there's, there's still very specific engineering disciplines that you know are going to be strong. Chemical engineering, I can tell you it is a great time to be a chemical engineer. Because there's so many new processes being invented now for sustainability, for recycling of plastics or figuring out how to capture CO2 or I mean, for a period there for chemical engineers, mostly what you did was refine a lot of existing processes. Now there's a lot of new processes being invented. Energy. If you're in the energy space, that's another huge one. Okay. That as we continue to figure out how to get to a no carbon or at least low carbon energy future. So I think there's some specific disciplines that are going to be very good. But no matter what you're in, I think it's also going to be more important than ever that you understand how the products are used, what the market's looking for, and have that knowledge, not just the specific skill of deployment. Another thing I tell every engineer. You need to learn about finance. Maybe you don't care, but the bottom line is in a public company that is a for profit company, decisions are driven by financials from, you know, eventually they get to a decision that directly affects you, but it always starts up here somewhere in a decision in what area to invest in or. Or what markets we're getting to or not get into. And you need to understand how that drives your company. [00:39:53] Speaker B: Perfect. Peter, it was a pleasure hosting you on this podcast and really appreciate your time and appreciate the candid with you to share your thoughts. [00:40:01] Speaker C: Okay, thanks Vinny. Great talking to you. [00:40:08] Speaker A: Thanks for listening to INA Insights. Please visit INA AI for more podcast publications and events on developments shaping the industrial and industrial technology sector.

Other Episodes

Episode 80

May 11, 2026 • 00:39:57
Episode Cover

Prith Banerjee: Synopsys's Bold Vision for the Future of Engineering

Digital twins can already predict the failure of a million-dollar industrial asset with 99% accuracy — so why are engineers on the ground still...

Listen

Episode 8

February 28, 2022 • 00:37:09
Episode Cover

Victor Mendelson: HEICO – Troubled Asset to a “Hotshot Stock”

HEICO Corporation’s turnaround story has created headlines for the past 20 years. The company designs, manufactures, and sells aerospace, defense, and electronics-related products and...

Listen

Episode 16

February 14, 2023 • 00:26:06
Episode Cover

Omar Asali: Driving Innovation & Sustainability in Industrial Technology

Omar Asali, Chairman and CEO of Ranpak, shares his viewpoints on the increasingly transformative role of technology in the industrial space. As head of...

Listen