Industrial AI for the Physical World: Siemens’s Peter Koerte

Consumer AI makes headlines daily, but industrial AI increasingly enhances and enables nearly everything we do. Learn how one multinational company approaches data management and deployments at scale on today's episode.

I'm Peter Koerte from Siemens, and you're listening to Me, Myself, and AI.

Welcome to Me, Myself and AI, a podcast from MIT Sloan Management Review, exploring the future of artificial intelligence. I'm Sam Ransbotham, professor of analytics at Boston College. I've been researching data analytics and AI at MIT SMR since 2014, with research articles, annual industry reports, case studies, and now thirteen seasons of podcast episodes. In each episode, corporate leaders, cutting-edge researchers, and AI policymakers join us to break down what separates AI hype from AI success. Today we're talking with Peter Koerte, Chief Technology Officer at Siemens.

Siemens is a German multinational technology company focused on industrial automation, smart infrastructure, and mobility systems, all increasingly important topics. We'll discuss industrial AI, what it means for the workforce, and what the implications are for data sharing across industry. Peter, welcome.

Oh, thank you, Sam, for having me.

Great, let's start at a high level. Some of our listeners may not be familiar with Siemens. Can you give us a brief overview?

Yeah, sure. So Siemens is out there since 180 years, almost, and uh, what we say is we transform the everyday of everyone. And what that means is, if you think about the chair right now that you're sitting on, the clothes that you're wearing, the water that you're drinking, the electricity that you're using, the transportation system, such as trains that you're using on an everyday, all of that was enabled by Siemens when it came down to the way we design these things, we produce them, how we actually make sure electricity is safe and distributed, how transportation is run smoothly and safely.

All of that is coming through Siemens, except as a consumer, usually you don't see us. But in the industrial world, Siemens is a very, very big brand name, and uh, we are recognized for the high quality, but also for the great solutions we bring and the simplicity to our customers.

Yeah, I think that's a great example, cause so much of the world we rely on, we just don't pay attention to. We don't notice it unless it isn't working for some reason. You talked about industrial AI. What exactly is the difference between industrial AI and consumer AI that most people would would be familiar with?

Yeah, the big difference is today, of course, consumer AI is making the headlines, while we think industrial AI is quietly but profoundly, uh, changing the physical infrastructure, the physical world that we know of. So think about, for example, the building that we are sitting in right now, and that building has, of course, some uh, climate control. About thirty to 40% of all the electricity that we're using today goes into buildings. What we are saying is what if we actually can take all the sensors that we have in these buildings, then actually develop an AI that automatically learns every minute or fifteen minutes in that case, and then automatically adjusts all the temperature settings, all the lighting settings and everything

in order to cut cost and energy. And that's exactly what we're doing. We just launched an application that saves 30% of your energy bill, and therefore reduces greenhouse gases by 30%. Just by doing that, it runs autonomously in the background. And this is what we do for grids, we do this for factories, we do this for machines, we do this for, of course, buildings, and we do this for trains. So everything in the real world, we are making it more efficient simply by what we say, connecting the real world and the digital world, where we try to optimize and make things better.

Yeah, that makes a lot of sense. I mean, I'm sitting here on a university campus. It's spring break, and I guess we are probably heating this place about the same as we would be if it was full of people. I don't, I don't even wanna ask. I don't wanna know here. That's it. That's But I think we're all familiar with consumer applications, and I think the failures of AI in consumer applications get a lot of attention, you know, with the hallucinations and these sorts of things.

Somehow that seems very different if you're connecting this to the physical world. It's not just a funny anecdote that goes across the internet when AI screws up. It could have some real world consequences when you make that connection. How is Siemens thinking about that?

Yeah, you're absolutely right, and and Sam, thank you for saying that. So when we compare consumer AI to industrial AI, there's three things at the very least that are profoundly different. And the first one you already alluded to is, uh, the level of precision and accuracy of those models. So obviously, you don't need any hallucination when you make recommendations for an engineer to design the next part for, let's say, your smartphone. Or, uh, you certainly don't need an AI mistake when you think about how to optimize an electricity grid because that's critical infrastructure.

So what we need to ensure is the highest level of quality of those models, which, as you can imagine, that's where we get to into ninety-nine, 99.9, and so on, the percentage of accuracy of the models. And a lot of work goes into that to make sure that these are reliable, safe, and trustworthy. That's the first part. The second part is actually, how do you train these models? Because I think we, you and and all of us, we are very familiar with what we call large language models.

Now, in industry, we not necessarily talk about large language models. We usually talk specific data when it comes to going back to the building example, temperature settings. So we have a lot of time series data. We have construction data, we have engineering data, we have simulation data. This is very different. So these are geometries, pictures, vectors, what have you. And so we have to make available these models in a very, very different way. And the third difference is, how do we get that data?

Because when we build these models for the physical world, we cannot just go on the internet and just download a bunch of data from sensors for your buildings or CAD data or whatever. This is very often even very proprietary data. So customers are only willing to share that data if we are able to express an incremental benefit of when they use our model, then they in return share the data with us. So, of course, in your case, better energy savings in the building, but and also for designers, faster time to market because we can get them designed faster and so on.

So that way of how you actually get to the data is very different. So the language you're speaking, the accuracy that we need, how we get the data, this is in the industrial world quite different than to, of course, what we use in the consumer AI every day.

That's pretty fascinating. My naive reaction when you first started talking was that, oh, you know, what you're describing is much more structured data. So I was pretty excited when you were talking about, oh man, all of this data is temperature data or structured data. But the idiosyncratic nature and how it applies only to your building or only to your machine and only to your setting seems very difficult. Tell us a little bit about how you're getting people to give that data to train machines and how that transfer works between organizations.

Yeah, it's a very good question, and you're absolutely right, because if you think about it, if I say it's a great day, or the day is great, the LLM does understand the meaning that actually it's a great day. In engineering terms, it's very different. So therefore, we need to adjust and cater for that. The way this works in the industrial settings is you go, of course, after the industries step by step and say, okay, what is the semantics in there?

I I alluded to buildings, and in buildings there's certain standards and there's certain data formats and what we call ontologies, so it's the semantics. And there we try to get that understanding about what is it that data that is there. It is more structured as you say, but as you can imagine, right now you're sitting in a room with Fahrenheit, I'm sitting in a room with Celsius. And so therefore, if you then say, well, even this is a temperature setting, actually, it's it is, but it's quite different.

If I'm talking twenty and you're talking twenty, right? So for me, it's warm, and for you, it's actually really freezing cold. And so that that's something to adjust for. So it's it's not a slam dunk, but understanding these use cases by industry by industry is is really key. And so in buildings, it's all about energy consumption. But as I said in engineering, very often it's time to market, it's in production, it's usually quality and throughput. And so understanding the data and the the key variables that drive that is important, which brings us to a key word that I want to mention, and that is called domain know-how.

Because you can argue, well, any data scientist can do that, and it's true. However, you really need to understand the domain that you're operating in and the key parameters. And I'll give you just one very simple example, but I find it fascinating. I'm not sure when you used the last time a train, but maybe the next time you use a train. And I ask you, what is the most critical component of a train? And probably you would say, well, probably the brakes, and that's true, it's safety critical, but it turns out it's the doors.

And why is that? Because if you think of the job to be done of a train, is to move people from A to B. That means it stops, it gets people on and off, and so you go from station to station to station. So the whole day, indeed, yes, the doors of the train go open and shut, and thereby they break down. So the most critical part in that regard for an operations is the door, and this is domain knowledge you need to understand that part.

Once you understand that, then it's fascinating because then what you can do is you can say, well, give me the voltage reading of that motor that drives the door. Look at, of course, the profile of how that motor operates. And in the meanwhile, today, our models can predict any door failure ten days prior of its failures. So therefore, you can get it into the depot and you can fix it, which means higher uptime, higher reliability and all of it, and better passenger comfort.

So these are the examples where you have to combine the domain know-how together with the technical know-how, meaning AI, and that's how you create customer value, industry by industry.

I like that because I can get my mind around that example. Some of the things that, uh, that I was reading about Siemens were complicated to understand, but that makes a lot of sense. I think everyone has some sort of application where they would like to know ahead of time that something is going to break, and before it breaks, because when it does, it's a mess. Siemens doesn't necessarily own trains, though. So how do you get that data about those voltage into your systems versus your customer who has purchased and bought that train?

They have to have some sort of way to send that data. They've got to share information with you somehow. Weirdly enough, they would benefit from someone else's train data for a train they don't own. How do you manage that infrastructure?

Yeah, it's a great question, and it is that's why I said it's very different the way you collect data in the industrial world. And let's stay in the train example. Truth be told, those customers, they just simply don't. They just say, give me the train and I'm fine, and then I build my own model. And so we have operators like that that do so. Usually, however, they are not the ones that are most successful. Usually, the ones that are saying, you know, you can send me a train, but the train, if you look at the total cost of ownership across the entire lifecycle of a train, which is, let's say, thirty years, the train in terms of capex, so the

investment, is about 10% of the TCO, 90% is operations. So what if I go to you as the OEM, you know best your system, and I share the data with you and you help me to optimize? So you help me to optimize with regards to reliability, that's the dry example. You help me on the efficiency. This really goes down to, of course, the way you operate that train. So we have believe it or not, we have AI that helps you to think about how to accelerate and decelerate or break that train in order to save energy.

So energy is one of the biggest, you know, operating costs that you have on that train. This is where we take then that data, it's connected, so all of these devices are then connected and then, of course, reliably and encrypted. And then we have that data and then we make use out of this data and we build our own models in that regard. And we do this customer by customer, and very often we do have data sharing agreements so we can use the data.

We don't own the data. It's important. It's still our customers data, but we can use it and train our models for their purposes. And then, as you said, we can combine it with other data, so everybody gets better in that regard. And that's exactly what's happening, not just in, let's say, in trains, but you see this in many machines. But it turns out it's not enough data to build your own models because you need to have much more, much, much more data across different settings.

And this is where Siemens comes into play because usually we don't build machines and we don't build all the trains, usually we build components that go into it. So so we work with car manufacturers, we work with, uh, aerospace manufacturers, we work with life science companies, we work with food and beverage companies, and so on, in order to help enable them. And so they come to Siemens and naturally say, you know, how can you help our specific industry to become better?

Hadn't quite thought about it that way that if one person has insufficient data to train a model by themselves and another person has insufficient data to train a model, but together they do, then the idea of connecting those people together creates value that neither of them could. And we had a guest from Land O'Lakes on a prior episode. They're sharing information with farmers, so farmers build things, they have a lot of data about their crops, but how they share that data.

I feel like there's a lot of that going on where we're recognizing that idiosyncratic data is more valuable when combined with other data. At the same time, I'm not sort of naive. People don't want to share stuff. How do you encourage people to do this?

There's a simple not an easy but a simple answer to this, and that is value. So if I'm not able to translate that and say, you know what, share the data with me, uh, and then thereby you kind of improve your reliability of the train to stay there, or I improve the efficiency of your building, then they will not share the data. It's as simple as that. But if you do, then that's great, and then they say, that's fine.

And sometimes it's built into your solution, so there's it's built in at the contract where they just say, well, we don't care, it's fine, you can just use it. Others are saying, hey, I want to have also, you know, a negotiated discount, which is also possible. But the simple answer is you only share your data if you get some value in return. So that's a little bit like the model, and it's just an industry. It's slightly different in terms of the kind of value we're creating, but still, there's some value in return.

You're describing largely a partnerships, but sort of between customers or with customers. But you've also done some recent connections with industry, and like, for example, your partnership with NVIDIA. Can you describe what you're thinking there and what an I think the goal there is an industrial operating system. How does that work? What's the plan there? What's the thinking?

Yeah, so with NVIDIA, uh, we have a very, very close relationship because for many reasons. Uh, one, of course, we do lose a lot of GPUs in order to train some of our models. A second tool that we're providing today, Siemens is the leader in industrial software, so we are about 10 billion euros of digital sales. People forget about that. So so we are among the top twenty of software companies in the world. So we have a lot of simulation software where you can simulate cars, trains, rockets in the digital world.

And, of course, all these simulations take an awful long time when you think about computational fluid dynamics or and so on, which is very complex. But it turns out you really can accelerate them. And so what we're doing together with NVIDIA is to say, what if we actually, instead of waiting eight hours for a complex computational fluid dynamic simulation, let's say of the air drag on a car, instead of eight hours, we can reduce that to minutes.

And that's exactly what we're looking at. So it's accelerating simulation, accelerating design when it comes to chip design, for example, in design, which is really interesting. As we get to lower, lower nanometers, two nanometers and and less, the complexity of verifying those chip designs is enormous. It exponentially really rises. So instead of having human engineers going through every, every circuit and really testing it to every gate array, actually, you can start to have an AI go through this and do this over and over and over again.

So the chip design verification is one. And then lastly, there then the design transfer to manufacturing is a key issue because every of these really holds you up in how fast you can get these chips out there. And so there again, we can already, as you are the designer, we can have the AI in the background verify whether what you have designed is correct and whether it can be manufactured. And these are kind of examples that we have announced also at CES earlier this year with NVIDIA.

We are really excited about because we think we can further accelerate, and this is always the key word, acceleration of design, acceleration of manufacturing, acceleration of operations, and that's why we are so excited about it.

I get the appeal of switching eight months to eight minutes. I mean, that Yes, it's amazing, yeah, clearly,

isn't it? Yes.

clearly, pretty yeah. I, uh, it doesn't take much quantification. We can do that in Fahrenheit or in Celsius, either way that works. But the other thing it makes me think about is that you probably have a lot of processes designed around the idea that it was going to take eight months to do that. And when it takes eight minutes, it feels like sure, it compresses it, but it also might change the types of things you do, the order that you do them in.

It seemed like it could just have this ripple of upheaval. How do you manage that? Or, you know, am I extrapolating too much? It? It feels like it could be a mess.

that is it's very true. That's why I tend to say always AI is about 20% is technology and 80% is actually transformation. And what that means is we talked a lot about data, that's one thing, but then it is really changing the processes of how you do things. And usually what the AI is now doing is it really changes workflows. So instead of thinking sequentially, where, you know, I do one task, let's say I do the design, the next one is doing the verification, the one next one is looking at how do I design to transfer, transfer it to to manufacturing, it's very sequential.

Now, what if you could do this all in one step because the AI is doing it? So obviously, you're disrupting a very well established workflow process. The first question that comes is, who is doing this? Is that the designer from the very end, from the beginning? Is it somebody completely else? So who is the persona that you're actually talking to? Some very interesting questions. Second, how is that process then gonna go? And who's verifying that whatever the AI is doing is is really correct?

Then a third question is, where do I do this? Where is the AI sitting? Is that a new application? Is that embedded into an existing application? Is it talking to all applications? All of these interesting questions arise, and they are not usually all technical. Very often we find this is very much about the people that use it every day and involve them and then start to think, rethink what wasn't possible before, and of course, thereby addressing also some anxieties.

Because many would then argue, well, the AI is gonna take my job away. So then you have a lot of resistance, and then all of a sudden a technology conversation becomes into a cultural, cultural change, transformation conversation, and we find this time and again.

No, the natural follow-up is for me to ask about workflow and these type of issues and they're all important, and I don't want to discount those or whatever. But you're pretty fired up about smart glasses and workers wearing smart glasses. What's next for them? How do you see them in the industrial world?

Indeed, I'm very excited about AR glasses. If you think about, in particular, US. manufacturing, I just spoke to a major new EV electrical vehicle car manufacturer, and they told me in their manufacturing, their churn rate, so the attrition of their blue collar workers, is 35%. What that means is you constantly have to retrain your employees, and it's it's not just retraining them, but also the other question is, how do you capture that knowledge? So what if you can take your glasses, you have that camera, and let's say you are a specialist in, uh, operations, and you are a maintenance engineer for a specific machine.

That actually, camera and that AI is overlooking your shoulders and literally and really is checking off what you're doing. Maybe you're narrating it even, you record this, you do this over and over again, thereby you're democratizing that knowledge. Actually, you can capture this. So for future people coming in, but even better than for the new worker, working the night shift, two a.m. in the morning, a machine breaks down. Usually, people are just tinkering around, having no idea.

But what if you had those glasses on now? And it's saying, those glasses are saying, you know, this is a CNC machine. Usually, the failure code of E three, four, five means, uh, actually, it is a jam tool. Check a lead lead and, uh, open this one, two, three, four, five, and off you are. How amazing is that? And so I really think in terms of the keywords augmentation, so augmenting the workers, the blue collar workers, but also white collar workers on the shop floor, and, of course, capturing that knowledge as they're exiting.

Isn't that amazing? And I think it's gonna make us all much more productive and much more enjoyable because, you know, you get faster time to result, and thereby you get the factory running and so on and so on. And you reduce a lot of anxiety and fear because very often people don't know what to do. And now, all of a sudden, they have a companion, they have a co-pilot colleague, whatever you want to call it, that helps them, and that is there for them twenty four seven, as opposed to calling somebody who's probably somewhere home and sleeping.

Hm, that makes a lot of sense. I want to draw a little contrast, though. Earlier, we were talking about data, and you were talking about a need for deep expertise and deep domain knowledge. But it sounds like this is maybe a push against, or, you know, you're not needing to know that the E three forty-five error code means this, that, or the other. Is it deeper? Is it more specialized? Those seem in conflict to me in some ways.

Well, obviously, we need both, but it's actually the example is is pretty comparable if you think of it. So, yes, I can tell you the door is gonna break down, and this is now preventative maintenance. The other case was more as a reaction to, but in both cases, it's maintenance. And so the preventative maintenance means that still a worker has to go out there and replace the motor. Now, on the other hand, in our case here, it's the same thing.

It just gives you the intelligence of what to do, and the doing itself still has to be done by somebody who is operating that machine. So I think it's pretty comparable. The interesting thing about this is because it still requires humans, could we at some point automate that through the whole conversation about robotics and humanoids and everything? This is certainly then also a big push right now that we're seeing in the market, uh, whether this is gonna come soon or not, we don't know, but for sure, we're missing at least 2 million people in the workforce in the United States already today, uh, to get them on the shop floor.

So we need the only way how to stay productive is by automation. And so this is where Siemens helps also many companies to automate their processes in the factories.

Maybe I'm reading too much into it, but I read something you'd written about humanoid robots and some skepticism about the actual humanoid shape, and you were kind of hinting at that right there. For one, I'm totally with you. We the human shape is not anything magical, and there are a lot better shapes for industrial machinery in particular. Are things gonna look human, or are they gonna look like machines, or different?

Well, that's the big debate. To be honest, it's too early to tell. I've seen both, and matter of fact, today I just had two conversations of that sort. One of was, uh, them going into the direction, we need to have humanoids. The other one was actually saying, no, no, no. I think in the end, it comes down to, uh, the ROI and the value again that we're creating. Let's take a very simple example. Um, let's say material handling is a is a big one.

Uh, in a factory, you have to always make sure that there's ample supply of material. Let's say, in particular, if you're in a stamping plant and it's metal sheets and so it's heavy. Taking a humanoid is probably not a good idea, although there's use cases, I've seen them, and there's many reasons. One, the payload is very, very, very limited. Uh, number two, humanoids are quite slow if you look at them, at least today. The question is, can you accelerate them?

But today they are slow. And then, lastly, is up to 30% of the energy consumed in a humanoid is just to making sure that you're standing upright. And what if you actually had different form factors that would give you higher payload, faster speed, less energy consumed, and then it becomes an ROI conversation. It depends. It's very hard to generalize. In this case, though, I almost would bet probably a different form factor to a humanoid is a better one.

But there's others where you could argue a humanoid could do a better job. For example, wiring harnesses, clipping them together, where you need to lot of amb- dexterity and and versatility and all of it. Maybe, but that's exactly that's why it's a fascinating field. I think anybody who claims who knows it, um, I think it's too premature, but it's a fascinating field.

Actually, I like that because I think so many things are increasingly it depends that we don't have these one size fit all models that are going to work. And you know, that that defeats our ability to make some sort of prognostications here. Thanks for taking the time to talk with us and sharing your insights about industrial AI, which is probably a different idea for some people, and but also data sharing in the future of work. And listeners, thanks for joining us on me, myself and AI.

Thank you, Sam, it was great.

Thanks again for listening today. Next time, Venkat Kozla, CTO at the Washington Post, joins us for a talk about AI innovation in publishing. Please join us then.

Thanks for listening to Me, Myself and AI. Our show is able to continue in large part due to listener support. Your streams and downloads make a big difference. If you have a moment, please consider leaving us an Apple Podcast review or a rating on Spotify, and share our show with others you think might find it interesting and helpful.