Podcast #199 Industrial Data

TRANSCRIPT

Ken: Good day, and welcome to episode 199 of our Momenta Digital Thread podcast series. Today, I'm pleased to host Marcel van Helten, President of Red Lion Controls. Marcel is a passionate executive with over 25 years of leading strategy, planning, and execution of technology businesses across industrial software, Edge computing, the Internet of Things, embedded technology, automation, and digital health care. He has built and managed successful global commercial marketing and technology teams for GE, Kontron, and B&K Vibro across Europe and China. Marcel, welcome to our Digital Thread podcast.

[00:01:19]

Marcel: Good morning, Ken. I'm really glad to be on your show. Happy New Year, it's still allowed on the 9th of January, so I'm looking forward to it.

[00:01:27]

Ken: Absolutely; this is the second one we're recording in the new year. Happy New Year to you as well. I've been told it's okay to say it up to the end of the month because you have to capture Lunar New Year, which happens in about three weeks.

[00:01:42]

Marcel: Another supply chain disruption.

[00:01:44:]

Ken: There you go, baby. Look, you've come highly recommended for this podcast by our own Luke Smaul, who joined us recently as a strategy partner. I understand you worked at GE together, so I'm sure there'll be some interesting stories along the way. We call this the Digital Thread podcast, and we look at it as one's individual or professional digital thread that led one to the current position. What would you consider to be your digital thread?

[00:02:12]

Marcel: It's the digitization of the manufacturing space, what we now call the IT/OT convergence. I started about 30 years ago, selling SCADA and HMI systems in the early days. At that time, we had already combined manufacturing industrial data from PLCs into your HMI with your base data to optimize the output from the plant. We started the European business segment called Factory IT about 20 years ago at GE. It was about deploying software to run manufacturing plants more efficiently. My main digital thread is how we can use information combined with great technology to improve customer outcomes. It's focused on what outcome you can provide using technology and information and bringing that together. If you call that IT/OT convergence or whatever name we give it right now, using that is my digital thread.

[00:03:21]

Ken: It's interesting; you and I share a similar one. I was lucky enough to be part of a company called Wonderware way, way back in the day before it was public. It's very much about that intersection of IT and OT, so I've always had a natural love for the convergence of those two. GE certainly did it well, including this idea of outcome-based- both selling and solution-providing. You were 15 years plus with GE's industrial business in Europe, including Managing Director for Germany at one point. What were some notable learnings relative to industrial controls during that time?

[00:03:58]

Marcel: It was an interesting journey. We were a US supplier, almost mid-tier below the big ones, Siemens and Rockwell. A couple of things I learned- I think the industry is conservative, and I think that is still the fact that I remember when it started. My boss then told me every machine would have a telephone number. Now we're at a place where every new machine might have an IP address. We're already at that time, a lot of technology available. Maybe not as fast as we have now, and we couldn't probably handle the data capacity we have now. Still, we had a lot of technology available. What you see is the industry is very conservative in adopting new technology also because of the lifecycle of the assets, the financial lifecycle of the assets. It's very difficult or costly to upgrade control systems while the asset still performs to the customer's satisfaction. That is one thing; I think it's a very conservative industry, and we still see that. The other thing is domain expertise is important and understanding the industry and the customers is key to success. You need to understand what parts of your equipment are crucial to the operation and what parts you can try new technology on. I gave you an example, we were once doing a system for a big bottling line at a big brewery, and we were collecting all the manufacturing data to run all the efficiency reports and where we're using a commercial data off the shelf database, one of the larger suppliers using the ODBC interface. The system stopped every week. When we found out what it was, it was a memory leak into the ODBC interface.

When we returned to the ODBC interface supplier, it said, "You know what, why don't you restart the thing every day?" Correct? If you're in an office environment, that you could do. That is trying to understand how this is used, and that's the same thing- if you look at safety, critical applications, your safety critical. Understanding that and how things work is a lot different than creating just an app on an iPhone over the weekend. Correct? Some pieces are- you can do it that way, but some pieces are so critical they need to be very solid in how you develop the technology. Then the other interesting thing- so, I've been away from industrial automation at its core and back now at Red Lion. But the interesting thing, the fundamentals of industrial automation didn't change. We have new field buses, Profinet is there although when I was there- we had more standards some 30 years ago. Like, it's S95, and it's S88 for batch. But if you still look at the support of ladder logic in IEC 1131, that's old-fashioned. A lot of the fundamentals of industrial automation- although we improved a lot on the data information side, the fundamentals on the plant floor are still intact. That's an interesting thing to see, and that has to do- because at the end of the day, what we need to do is still need to fill a bottle or do an aerator of water treatment plants or those things. That's what I have seen over the 15 years at GE going away from industrial automation and coming back right now.

[00:07:37]

Ken: I know you went on to play leading roles at Kontron, Abaco Systems, and finally, B&K Vibro, all industrial system leaders in Europe. How did this time prepare you for your role at Red Lion, which we'll talk about in a moment?

[00:07:52]

Marcel: These are all technology companies, and it's exciting to work with a team, including the customer, which is always very important, who have ideas on what they want to do, and design something great which delivers customer value. That's what we tried to do. For example, we worked with the medical industry at Kontron to help customers run their image processing in the Cloud. That was an interesting thing to enable remote hospitals to still have that capability; you could do image processing from an MRI in the Cloud while they don't have that expensive equipment. At Abaco, we were looking to use artificial intelligence to help save lives on the battlefield. That was an exciting high-tech environment where money was not an objection. We could use the best technology available. Then B&K Vibro- we were looking at how we could make the field of vibration monitoring simple and easy to deploy. Vibration monitoring is a field of specialists. People are not in lab coats, but they're certified people, and we're trying to make that a lot easier by applying AI. That helped me think through how to set roadmaps and strategies to grow a company and the interesting pieces of a business where you can grow because you can deliver better customer outcomes by using technology- that piece I still use every day. If you look at what we try to do at Red Lion, our vision is to empower our customers to unlock the value of industrial data, building a better world together. Building a better world together means being together with our customers. Red Lion started about 50 years ago in York, Pennsylvania.

Being in the industrial automation space, they bought a couple of companies. They bought Sixnet^®, N-Tron^®, and recently, we bought MB connect line, a remote access security company in Germany. If you started looking at a company from a distance, and normal things I tried to do is, what are we doing? What do we have in common in all those little pieces? What are we doing? It is about connecting your customer to industrial data. Now, either through remote access through our flag sets, IoT products, or the newly acquired MB connect line, or connecting assets through a network using our Ethernet network products in our internet product range or through an operator interface on the machine. It's how we connect a customer to their industrial data, and that's what we are focused on, and that is probably what I've done in those other companies looking at how it can help technology to create value for customers by looking at the bigger picture that helped me to prepare for the rollout at Red Lion.

[00:10:58]

Ken: One interesting thing is that you introduce those companies all as technology companies, which usually you think of as pieces and parts, if you will, from the technology solution stack. Yet, in every description you made of your time with them, you saw them much more in a solution or, to your point, outcome based. It's interesting; you always enter on the technology side, exit, or finish on the solution side. I can see that plan directly, even how you're approaching Red Lion to connect customers to their industrial data. I like that in your LinkedIn profile, you're using the byline, 'Building THE Industrial Data Company.' What do you consider to be industrial data? Where do you see the opportunity around it for Red Lion?

[00:11:47]

Marcel: Industrial data, the way I look at it, is generated by industrial assets. A lot is time series, correct? There were transactional as well, but the majority is time series data, and we've been working on time series data for a long, long time. What is interesting right now- I mentioned to somebody else the other day that industrial automation is cool again, and it's that companies invest billions in software to optimize their operation. When we talked about the Cloud Guy, AWS, and Azure-based applications, you see a lot of development and investment in that space. The biggest challenge remains access to the right data, which we are trying to solve. You need access to the right data to create value out of those systems. That is not just the PLC you can connect. It could be a blower, which is a standalone system somewhere, which has an impact on the product quality. Because if the blower fails, the temperature goes too high in the room, and the process changes. Unless you have that blower data, if it is running or not running as a part of your data analytics, you cannot do the right analytics, or you might come to the wrong conclusion. That's why it is so important that we provide companies that create solutions using all that information and the right data. What is interesting as well is that a lot of people working in that space right now don't have the expertise where we talked earlier about the plant floor. They expect to work with an information point in context, and they don't understand that that might be an old 4 to 20-milliamp sensor or PT 100 sitting somewhere. Bridging that gap and making sure that the people who do cool things using AI and all sorts of other technologies, building great solutions, make sure that they have the right data and can get to all the information from the plan floor, data is what we're saying, you know, empower our customers by unlocking the value of industrial data. That's how we see industrial data.

[00:14:15]

Ken: It seems like you're taking on a very important role at a very important time for Red Lion Controls as president for their business. I'm curious. What do you consider to be state-of-the-art industrial data management? Whom do you see as some of the best practitioners of it?

[00:14:31]

Marcel: I think every company already generates a lot of data. I think a couple of years ago- it was a long time ago. I think the automotive was pretty much on the leading edge. But just in time, integrating supply chain into the manufacturing plant in manufacturing operations, using data, all that sort of things. I'm not sure how far they develop now, but one of the newer industries where we see a lot of best practices is the wind turbine industry. Wind turbines are pretty expensive assets; they are critical that they run 24/7, and there are a lot of variables in there. What you're seeing is that wind turbine operators have a lot of advanced software technology for predictive maintenance. Because if you look offshore, you must have 12 months lead time to failure because you need to organize a ship to go there. You want to combine things. But also, they have connections straightaway with the energy cost and training; they need to look at energy forecasting, and a lot of weather data is involved in that. If you look at that particular industry- remote assets, hard to reach, there are not always great fiber optic connections to centralize data and a lot of interaction with different systems. I was very impressed with what they are doing, and they are probably on top of what is possible today. I would see that as the best practice there. It's interesting to see what those guys are doing.

[00:16:18]

Ken: I fully agree. I know GE had a play in that at the time. For one of your peer companies, Vestas- one of our earliest M&A assignments was to help them find an industrial analytics company to acquire. We ended up steering them and helping them acquire Utopus Insights in 2018, which had just spun out from IBM as the smarter energy team. It was a great deal. They got patents, they got PhDs, they got running a business with a pretty good set of exposure out in the industry. But I agree with you; all those cases you just talked about, and those capabilities came with that and were critical, especially for a large OEM who wanted to provide the same capabilities to other OEMs' equipment. That was the key.

[00:17:11]

Marcel: Exactly. They say they want to integrate third-party wind farms. I forgot to tell you, the other great thing they're doing with the data is they go back to the designs. They're looking at the data, the real-time data, and high-frequency data and start looking at how they can optimize the wind turbine's design, so it's a closed loop. What they're doing is an example for every OEM, and they do it with high-frequency data. It's interesting to see.

[00:17:45]

Ken: I fully agree. One of the spiffs that anybody who refers a podcast speaker in Momenta is that they also get to suggest a question. Luke Smaul suggested this one. He said, 'Look, to be successful in industrial data means considering the Cloud and perhaps cloud vendors, such as Azure and AWS. How do you see this playing against the traditional Purdue model or industrial automation stack as we think about it, specifically around industrial data?'

[00:18:15]

Marcel: I think Luke knows my opinion about that. Yes, I'm not a big fan of those models. It's probably 25 years ago when we said we need to be user-centric, or maybe you should say data consumer centric right now. If you look at a typical model and the different layers, what I always found is if you look at the typical automation pyramid- ERP is pretty small, so the way we automation guys used to draw the world, they say, we will design all the POCs all the IO, and then we will draw a little wall, a cloud and say, ERP. Looking at the IT world, they will draw all those processes in the ERP world, and then a little line says plant floor. The fact that we think that ERP on the top is smaller is a misunderstanding. Then what we're trying to say as well is one of the things is- at the plant floor, we talk about milliseconds; on HMI, we talk about seconds, MES maybe we talk about minutes, longer minutes and hours, and ERP, we talk about days in time for data accuracy. It's completely vanished. We need to provide everybody with the right data at the right time when you need it. I will give you an example. We currently have a customer. They take a high-speed signal to one of our FlexEdge^® Cloud controllers we call the Edge controller. It drives data into the Cloud, using all sorts of other data they combine and determine in a real-time algorithm. They claim it's the best in the world to do that, a new set point for a drive. They're sending it out, the setpoint, and adjusting the drive to get a product flow. They say with this algorithm, they are the best in the world and the most efficient in operating that asset. It's only possible by providing real-time, high-speed information into a Cloud and getting it back in real-time.

Latency, for example, is a key issue in this whole thing. In a typical Purdue model, where you have to go to different layers or through the automation pyramid, that will not be possible because you will be on layaway for days. That is what we need to do, and our goal is to see that we can make industrial data available everywhere in a secure way. That's a new thing. What comes in now- of course, security is a much bigger topic than it was a couple of years pre-Stuxnet, and that is what it is. I think the model was flattened. You need to look at more time in the data bus and with different consumers. You want to put the consumer- it could be an operator, or it could be an analyst, it could be a piece of software in the middle, and you need to make sure that it has access to the right data at the right time when they need it securely. That's the way it looks. In that respect, the model itself is not doing justice to how the world works today. How do you want the world to work today?

[00:21:29]

Ken: I always drew an analogy between the industrial automation pyramid and Maslow's hierarchy of needs, which also has a pyramid framework and the idea, of course, being the basics that we need to be safe and secure at the lower levels and self-actualization at the peak. I think of that as an outcome framework. I do basic sensing or reading at the bottom; I'm developing deep insights, hopefully at the top. The unfortunate part of any pyramid is that the size should represent the importance and doesn't necessarily in that regard.

[00:22:05]

Marcel: Exactly, so how it goes back- it's a lot more closed loop, and it's just not up and down. It is almost chaotic how data moves between things right now. Some pieces of data will go depending on the application- classical, center, PLC, HMI, or classical DCS. Then going to the historical database, whatever process database, and then into the ERP or put an MES in the middle. But there is also data, which goes straight from the center into an application running in the Cloud, so it is much more chaotic now. I'm not sure chaotic is the right word, but there are many possibilities. We cannot just streamline in layers. Not anymore.

[00:22:59]

Ken: It's certainly apropos. What I appreciate in your background is you have depth in Europe but have worked at US companies and, of course, now, once again, in the leadership role of a US company. There have been some very bullish forecasts on manufacturing in the US, and Europe aligned, of course, with the CHIPS act and resilience reshoring efforts. How are you thinking about these markets and your business opportunities in 2023?

[00:23:24]

Marcel: We all know that some people will say the markets will slow down depending on which analysts you listen to towards the end of 2023 or early 2023. Unfortunately, we already have seen many tech companies acting because they over hired during the pandemic. I'm bullish about the American market. As you stated, the whole re-shoring, people want to get the supply chain back into the US, and we are looking at that as well. The whole chip supply chain crisis, including harbors in China, because COVID put a lot of stress on that. We see a lot of improvement, but I continue to be very bullish on the US market. OpEx spending will go down a little bit, and people will be more careful with maintenance costs and maintenance spending. But CapEx cost- if you see what they're going to invest in, electric vehicles- I think the energy sector will be strong.

I'm bullish about it. Europe, I'm less confident. I'm more concerned about Europe. Of course, we have this terrible war continuing to go on. That's one piece for us as an American supplier which builds the majority in the US. We, of course, have the exchange rate at the moment that's against us. We have a strong dollar rate. We'll have to see how China picks up again. We know that traditionally- especially the German market- has a strong export focus on China. Let's see how that is going to be picked up. The German OEM market may have an uplift because the US is now looking for high-tech quality machines, which could be an uptick for the German market. I'm very bullish on America. I call it a concern about Europe because of both exchange rates and a slowing economy.

[00:25:19]

Ken: Look, you've already mentioned some of the potential challenges, including exchange rates, supply chain difficulties, etc., facing the US and European manufacturing business. I'm curious. How are you preparing Red Lion to help your clients address these issues?

[00:25:33]

Marcel: One of the interesting things, if you talk about the US market, is labor shortage. I will never expect that I will be in a position to compete for labor against manufacturing plants. I know what a person at Starbucks roasting plant close to our plant makes or the person at Harley Davidson, which we have in York, Pennsylvania, as well. The fact is, I know that there is a whole competition for labor in the market. I think there's still competition for tech talent. That is one of the concerns I have, also, the labor market. The US can only get the productivity they need because of the reshoring if they automate. Because automation- you have seen that in Germany, the efficiency of the German plants, and the highly automated status of those plants, allow Germany, a high-wage country, to compete against low-wage countries. The US needs to get into the same thing. With our technology, we still need to connect many older assets in the US; we need to get that data to drive those software platforms that can help optimize a manufacturing plant. That's where we can help the American manufacturing base. How to automate things, they can drive through the use of industrial data, can try things centralized so we can reduce in- for example, people need it for inspection of pipelines, or that we can automate help to automate plants more so we can use the people for high level of labor, and that is a way to avoid the labor crisis in the US. I think that's how we can help people.

[00:27:27]

Ken: I would fully agree there. Our listeners will know I'm a big fan of Andrew Obin over at BofA, who covers all industrial, publicly held companies in the US. He came out with his end-of-year report and forecasts double-digit CAGR growth through the next five years for the industrial automation sector. For software, more specifically, 15%. I tell you, I'm a firm believer. I am very bullish on both, and the time is right. As you said very well earlier, industrial automation is cool again.

[00:28:01]

Marcel: I agree. Automation is cool again.

[00:28:04]

Ken: Yes, very much. In closing, I always like to ask, where do you find your inspiration?

[00:28:10]

Marcel: I've read a lot about people who have accomplished great things in technology. I'm a technology nerd who also works in business. I recently read an intriguing story. "Amp it Up" by Snowflake's Dutch owner, Frank Slootman. Prior to that, he was the CEO of ServiceNow, which took over the entire service industry and data domain that Dell now owns. It was also interesting to read "No Rules" by Reed Hastings about how he runs Netflix about talent density, how he deals with how to improve talent density, and how to get the best out of technical people. Then there's the time when I'm running. To stay somewhat healthy, I try to cycle. But also read books about getting into those sports and how they prepare for crazy things like 100-mile Ultras and how to deal with that, or people like Ultra runner Scott Jurek. I'm interested in how people get the most out of themselves. How do you get the most out of a group of people, whether in technology or otherwise, and work as a team to achieve great results? I think it's interesting to read about how other people do it and learn and pick up some tips along the way. That's where I get my ideas. 

[00:29:39]

Ken: I especially appreciate the runner aspect. As a former marathoner myself, I know the discipline it takes to be able to run even a marathon. I say, even because when you talk about Ultras, there's much more to it, like triathlons. But what many people don't appreciate is that it's a good time to listen to books on tape. As we used to call it, right? It's audiobooks now. Great stuff. Marcel, thank you for sharing this time and insights with us today.

[00:30:09]

Marcel: You're welcome. I enjoyed it; I hope your listeners will enjoy it as well.

[00:30:13]

Ken: I think so. I appreciate Luke for putting us together. This has been Marcel van Helten, President of Red Lion Controls and consummate tech geek and runner- because that puts the book ends around who you are in the middle, and somewhere in the middle, it says outcomes. Thank you for listening, and please join us for the next episode of our Digital Thread podcast series. Thank you and have a great day. You've been listening to the Momenta Digital Thread podcast series. We hope you've enjoyed the discussion, and as always, we welcome your comments and suggestions. Please check our website at momenta.one for archived versions of podcasts, as well as resources to help with your digital industry journey. Thank you for listening.

[The End]