In partnership withSiemens
Building a better train doesn’t end with delivering the railcars. When Siemens was asked to improve train reliability, the company added sensors and built digital models that could predict the need for door maintenance 10 days before a door actually got stuck—allowing mechanics to prevent delays before they happened.
Peter Koerte, chief technology and strategy officer at Siemens, says the potential of the industrial metaverse doesn’t end there. “The minute you start to connect real-world operations with a digital simulation thereof,” he says, “you can enable a lot of new services you even hadn’t thought about in the beginning.” When the covid-19 pandemic struck, and transit usage plummeted, those same sensors were repurposed to monitor capacity and ridership.
The industrial metaverse will provide an interface between the real world and a digital world, built on simulations and digital models of complex human systems like machines, factories, or cities. Koerte names five building blocks that will help the industrial metaverse achieve its full potential: these detail-perfect models, termed “digital twins”; simulations based on realistic physics; tools for seamless virtual collaboration; the ability to build immersive and photorealistic environments; and computing power that allows real-time responsiveness. All these capabilities are developing, and their continual improvement will fuel industrial metaverse innovation.
The industrial metaverse will be a powerful accelerator for digital transformation. The power of simulation will allow designers and manufacturers to get things right in the digital world before committing physical resources.
And while interoperability and platform challenges will have to be overcome to bring the industrial metaverse to fruition, Koerte is enthusiastic about its revolutionary potential. The metaverse will speed progress toward sustainability goals, he says, by conserving physical resources, building carbon considerations into design processes, enabling more accurate accounting of emissions, and advancing digitalization. He explains, “there’s a lot of very real-world applications that can help us make this much, much better.”
This episode of Business Lab is produced in partnership with Siemens.
The emergent industrial metaverse, report, MIT Technology Review Insights, March 29, 2023
Laurel Ruma: From MIT Technology Review, I’m Laurel Ruma and this is Business Lab, the show that helps business leaders make sense of new technologies coming out of the lab and into the marketplace. Our topic today is the industrial metaverse, which, unlike the consumer version, is based on simulation and large-scale digital twins. These are familiar technologies to manufacturing and R&D efforts. These large-scale digital twins, representing whole manufacturing plants, cities, or other highly complex human systems, can provide an interface between the physical and digital world that can make the real world work better for its inhabitants. Two words for you: virtual possibilities.
My guest is Peter Koerte. Peter is the chief technology and strategy officer at Siemens. Welcome, Peter.
Peter Koerte: Oh, thank you, Laurel, for having me.
Laurel: You have a position at Siemens that requires a direct connection between technology and strategy. How does this help the company not just innovate internally, but also with customer relationships?
Peter: Yeah, you’re absolutely right. As a matter of fact, we believe that everything has to flow backwards from the customer, obviously. This is where we look at how technology can make a difference for the customers we serve today. We listen always very carefully to what they need, not necessarily what they want. We see a profound shift in that our customers are not asking for just products, but for solutions. They’re also not just looking for solutions, but they’re looking for solutions that help them across the entire life cycle. As a strategist, usually you just would look at markets and market numbers and everything that’s out there, but as a technologist, you can listen carefully and understand what they really need.
I can give you an example. Trains is a fascinating one. In the past it used to be that there’s a train operator and they just say, “You know what? We need a new train.” Then they tender it, and usually at Siemens we would’ve been designing these trains, building these trains, and then delivering to them, and that’s pretty much it. But the point is you’re missing out on the entire life cycle. A train usually is in operation 50, 60 years, so the majority of the value created by transporting passengers or freight is actually during this life cycle phase. And so, they ask us, “Hey, Siemens. You’re building these trains, but can you actually help us to be more punctual, more on time, be more reliable?” I mean, we know those trains usually don’t run on schedule. I mean, you know it in the U.S. We have the same issues in Germany. “Wouldn’t that be cool?”
We said, “Yes, actually we can help you with that, because we can actually build sensors into the trains so thereby we know in what conditions they are, so that we can service them, so that we…” For example, a door very often breaks on these trains. We know 10 days before they break that they actually need to be serviced. In that sense, these trains are much more up and running and are providing punctual services. That’s a profound shift in technology, where digitalization enables you to get into a completely new universe, the operations of trains. That is something that a strategist would’ve missed out on, because you would have to find the market while you’re building trains, but now with technology, you can see that you’re not just only building trains, but you help running them more efficiently.
Laurel: I love that example. I think it’s a really good one because it actually shows the value to the customer, which is the government that buys the train but then also to the rider of the train. You’re actually servicing your customer’s customer to help ensure those trains run on time and safely.
Peter: Exactly. The great thing about this is... This is another one about digitalization. The interesting one is that very often you start with one use case, in this case it was about making those trains being more on time, but then COVID hit and you know what happened. I mean, nobody actually went on mass transit, and fear of contagion, and so, therefore, the operators ask us, “Well, can you tell us how many people are riding on this train right away?” We said, “Sure. I mean, we have built in the sensors so we can look at the trains and see the capacity levels.” This is fascinating. The minute you start to connect real-world operations with a digital simulation thereof, you can enable a lot of new services you even hadn’t thought about in the beginning.
Laurel: On to the industrial metaverse, when we think of the industrial metaverse, it looks to create this bridge between the physical and digital spaces. Where is this technology now? What are some of those current use cases and what are aspects that are still in development that we can look toward in the future?
Peter: Yeah. Well, I really liked your intro. You said it well, there are these large-scale digital twins. This is precisely the way we look at it as well.
First off, there’s... Suffice it to say, there’s no clear definition of what the metaverse really is. There’s a lot of imagination in there, which is probably why people get so excited, but also so disappointed by it, because everybody is projecting their biggest hopes into it but then realizing, “Well, actually it’s old wine with new skins.” But in the case of the industrial metaverse, we look at it as something where all the building blocks exist today, they just don’t work perfectly together yet, but they are becoming more powerful.
Hear me out on this. We’ve got five building blocks that we think are important in the industrial metaverse. The first one is indeed the design. If you want to have a digital world next to the real world, we need to design or have a copy of the real world in the digital world, so that’s our digital twins. Then once we have that design, we also want to have it behave the same way in the digital world as it does in the real world. So, therefore, now you need to have simulation capabilities, so in terms of physics and thermodynamics and everything, it behaves quite the same. Then there’s the collaboration aspect, because you really want to bring people together. There is the photorealistic or the immersion aspect, so that it does feel real. The first instance is really making it look as if it is very real.
Then the last one is the real-time piece, because think about, let’s say, a complex car and think about 20 designers scattered around the world immersing themselves into the virtual world, seeing the car very clearly in a photorealistic way. But then one designer may say, “Well, what happens if I change, for example, the headlight, and the dimensions would be different?” Today you can’t do this in real time because the process and the compute power isn’t there yet in order to enable that. You can do some very small design tweaks, but not significant simulations in real time. But, the good news is, with increased compute power that is coming online, either on the cloud or on the edge, that becomes really powerful and so, therefore, the algorithm becomes smarter and becomes more and more and more real time, but still there’s a significant time gap.
What I’m trying to say, these five building blocks exist, except they all have to become more powerful and they have to become more connected.
Laurel: Earlier you were giving us this great example of how adding some digital capabilities to physical trains will help the trains themselves run better. When we think about the industrial metaverse and this idea of simulation, could you get into a little bit more of an example of, say, a car being built? Why is it important that you would have this hands-on ability in a virtual world that would affect outcomes in the real world?
Peter: Well, there’s many reasons. First off, there’s that you can get it right in the digital world before you actually build it. That’s actually how simulation, by the way, started. It started with cars because you always had to do the manual crash test, which turned out to be a big deal and it’s very costly. So this is why simulation started there. The whole thing is about being faster and having more iterations and having more people collaborate and integrate there.
Think about it. In the past it would’ve been just between, let’s say, the design engineers sitting together. But with this collaboration and real-time photorealistic rendering, you can lump in marketing departments that can give you feedback on that. You can lump in the manufacturing guys who can tell you, actually is that feasible, in terms of can you actually really build it. You can lump in the workers and see whether they can actually produce and assemble the car.
It really is all about time to market, if you like, number one. Number two is of course it’s really optimized, and then lastly, it’s also becoming more efficient because, as you can imagine, there’s so many requirements today. Cars are a great example, with regards to mileage and optimization of their energy efficiency, and so of course the more you can optimize them in the digital world, of course it will have a profound impact in the real world.
Laurel: Okay. With data coming in from so many different places, with all these iterations as well as different inputs, the market is also changing so quickly with consumer demands. So not only do you have these internal demands, because you can do many iterations, but you also have external demands on companies as well. How can the industrial metaverse help accelerate digital transformation for enterprises?
Peter: Yeah, that’s a very good question. Turns out that still this digital transformation piece is very complicated, isn’t it? We have been talking about this now for over a decade and it really takes a long time.
To understand that fully, let me put one concept out there. Before you actually become a fully digital company and transformed, you have to do three things. The first one is you have to digitize, so you have to get the things from the real world into the digital world. This is usually the step that takes the longest, because the return on investment is not all high, because you’re running the manual process and the digital process side by side. Very often that requires a lot of infrastructure that you have to put in place, new capabilities. So many companies that we see in the B2B world are struggling with this first step.
Then comes the second, what we call digitalization. That is bringing the different data silos all together, because usually companies are set up in silos, aren’t they? I mean, they are organized by regions, functions, business lines, or whatever. The power of digitalization is truly horizontal, bringing different topics together, different data points together. For example, in manufacturing you have maybe just the machine data and you optimize the machine data, but the minute you connect it to sales and understand what you need to deliver the next day, then it becomes really powerful. So breaking down data silos. That is really the power of digitalization, and this is where then it really speeds up and accelerates.
Then, lastly, the example that I gave you was the train. It enables you to change your digital transformation with regards to changing the business model. So instead of just selling your product, actually you have a lot of services attached to it and a much closer link to the customer. This is what we think the industrial metaverse is going to do, too. However, it’s predicated that you have to take this very first step. You have to have digital representations of your elements, of your assets that are existing in the real world. If you don’t have that, that’s really tough.
But there’s many industries, car manufacturing, pharmaceutical industries, food and beverage, media industries, that are really far ahead. They have these digital assets and so now are building these digital twins. For them, it’s relatively easy.
Laurel: With so much computing power required for digital twins and many of these industrial metaverse use cases, how can the metaverse be built in a sustainable manner? Because that is certainly something enterprises are looking toward digitalization to help with.
Peter: Yeah, I know. That is indeed a big question and we get that question a lot nowadays, particularly because of crypto and cryptocurrency and the whole notion of proof of work and the way it works. It’s very, very energy consuming, that’s true, but I think we are all in agreement that maybe that’s not the best value or time spent on making this work. In the case of the industrial metaverse, we believe that it is actually substantially helpful to have those simulations in the digital world first and then put them into the real world.
From the numbers I know, ICT (information and communication technologies) contributes about 4% to greenhouse gas emissions. Now, there’s other studies that I’m aware of that would suggest that up to 40% of greenhouse gas emissions can be reduced because of the digitalization effect, which we do think is possible. So there you go. It’s a factor of one to 10 in terms of leverage. So, yes, there is some element where you have to invest into it and probably create a little bit more greenhouse gases, but the net effect is absolutely very much in favor of doing it.
Last point on this one, the key thing today is that we need to understand the carbon footprint that we are leaving behind. Today that is a gross estimation. Today we say, “Well, around about 50 gigatons of CO2 equivalents are being emitted every year.” But that’s a simulation, that’s an estimation. We really don’t know. That’s not the real number, but you need to get to the real number. So what we’ve done is we created a low-energy blockchain, which uses as much energy as two clicks on a webpage. That enables you to communicate your product’s carbon footprint between different manufacturers, so that at the very end of the chain, you can actually sum up all the carbon footprint, based on true values, so that you know, for example, how much of a carbon footprint your smartphone produces. That is the step we need to take first, so the baselining of the carbon in the designs, before we are actually going then to the reduction.
Laurel: There’s also something to be said, too, that by having the digital twins and this industrial metaverse opportunity, then things like trains and cars and other large manufacturing facilities could be then made more sustainable themselves, because you are able to do it in this environment of simulation. Does that sound right?
Peter: Yeah, absolutely, absolutely. We tend to think about, if you like, the green digital twin. Think about a designer today. What does a designer do? The designer usually has a time schedule. You have to design this product by X. It must not cost more than Y, and it has to serve these functional properties, in terms of it has to go that fast or it has to be that stiff by Z. That’s the way it goes. We think there’s now a fourth dimension and that is the green aspect, so the green digital twin where you say, “And it must not exceed that many tons or kilograms of CO2.” This is where you have now an additional element of optimization that has to come into it. It’s a trade-off, isn’t it? That’s what is happening as we speak, and those calculation tools enable you to come to the best trade-off, as I said, before you even build these devices, buildings, factories, what have you.
Laurel: We’ve gone over some of the benefits of digitalization of industrial IoT (Internet of Things) in the industrial metaverse: data, time to market, responsiveness to customers, as well as this ability to improve sustainability, but what are some of the challenges? Why aren’t we all there yet?
Peter: Well, as always, there are many. First and foremost, there are of course the legacy systems. Every company has its own IT systems, its own configurations, so, therefore, most of the technology that we want to implement of course is not scaling the way it should and could. Second, very often there’s no interface, either from the machine where you can extract the data or from the software where the data resides. This whole notion of being open and being able to access other applications’ data is really a key obstacle. To me, to sum this all up, it’s really the whole question about interoperability.
We just recently launched what we call the Siemens Xcelerator, which is a digital business platform where we promote portfolio elements. So solutions that are truly open, where you have interfaces, so-called application programming interfaces (APIs), that are open. They’ll describe where others actually can build atop of it, and that are also very flexible so that you can install them in existing brownfield environments. That is really the biggest challenge in the industrial world.
Of course, as you can imagine, there’s always a human side to it, too, because I think it’s human to fear providing too much transparency. “Oh, boy, what happens if others can see what I do?” A lot of this is change management and bringing different departments, people together and showing them that actually this is not threatening them. As a matter of fact, it makes them better, because they can serve the customer better, and so therefore they stand to have less escalations, have more demand, and of course more collaboration with other departments as well.
Laurel: Yeah. About that change management, digital transformation really requires much more of a cultural transformation, doesn’t it? In your mind, even though we’ve been going through this process of digital transformation in many enterprises for many, many years, this new opportunity with the industrial metaverse, industrial IoT, and this physical-digital transformation opportunity, how do smart companies bring in people to be part of that transformation so they don’t feel so scared and left out?
Peter: Yeah. Well, I think that it’s one of the toughest questions. And there’s no simple answer in terms of a recipe or a playbook, because if there were, then it would be further along. But there’s a few patterns I would say are common.
Let me give you an example, because at Siemens we are also a manufacturer, which is great because we produce, for example, automation equipment, we produce motors, trains, and such. Just recently we opened a new factory, and that was completely digitally designed as a virtual twin or a digital twin. The way we did that was that we included all the respective departments from the very beginning and treated it as a change management project. Because the minute we built that plant or that factory in the digital world, we looked at not only how the structure would be, but then where should the machines be, how would the material flow, and everything like that.
The workers were involved in that, so they already knew what was coming. Even better, they had a say in, “Actually, aha, this is the way the machine operates.” Then all the other departments had their say into it, too, in terms of the safety department and whatnot, so that everybody could see what was about to be built and they were heard. You could certainly have gone faster, and just built it in the digital world and then put it in the real world, but the fact that they included so many different departments made them understand the bigger picture and was less threatening.
The greatest thing of all of that was that by the time we opened that new factory—that first of course existed in the digital world and then, after we had built it, existed in the real world—that factory was 20% more productive. It saved 5 million kilowatt-hours per year in terms of electricity, reduced 3,000 tons of CO2, and 6,000 cubic meters of water. It’s a triple win, really. You’re becoming more competitive, you’re becoming more sustainable, and your people feel more empowered because they had a say in that. Whoever is listening in, I can tell you it is worth that effort to go the extra mile in the very early days. Although it feels like you’re slowing down, you actually accelerate because by the time you build it, you don’t need to explain it anymore, because everybody is familiar with the concept.
Laurel: That’s a fantastic story, and obviously the numbers are the proof that you would need for that to just bring everyone along in the beginning. How do you envision the evolution of the industrial metaverse, and what trends and technologies are you excited about these days?
Peter: We are excited about many technologies and I could go on forever on this. But answering your first question, I think for sure this is an evolution, another revolution. Each of these building blocks I told you about, they become more powerful over time. We see this every day. There’s Moore’s law that enables us to get more transistors on the chips, so therefore making them more powerful, and so therefore becoming more real time or enabling more real-time applications. That’s for sure going to happen. That’s not going to stop, and that is a little bit more predictable because these parts we know.
I think the little bit of a wild card, that we don’t know as much, is the whole conversation that we nowadays have with regards to artificial intelligence. Simply because you can speed up the design when you think about generative design. In the past, of course, there’s a designer, there’s a human that designs, for example, an air duct. Usually they have straight lines and they look ordinary, as you and I would know. But interestingly enough, if you put an AI on it, and if you were to say, “Optimize it such that actually the airflow is optimized,” you would come up with very different geometries and shapes that really looked like an alien. They’re very unnatural. However, they are much, much more efficient in that sense.
There’s a lot like these examples that are happening. One other one is about photorealism. The nice thing about it is that you can create edge cases. Think about, you could simulate how it would look if you were to drive with your car through a volcano, or the ash rain that comes down. That’s an edge case you cannot really reproduce in the real world, but you can simulate it quite realistically in the digital world. And then you can take the data points, and feed it back into your algorithms, and see whether your car would be driving, just as an example for illustration purposes. This you can do for a factory as well. So AI for sure will have a huge role to play in that.
Then there is the question about the whole story about immersiveness. Today, the metaverse by and large is consumed through 2D screens, which is fine. But of course it would be much more powerful if you could experience it in 3D, and if you had your headsets or AR/VR headsets and such. But they’re still too heavy. They’re still too energy hungry. In terms of the frequency and repeat rates, you feel nauseated, all of that. As you know, a lot of money is going into this, and I’m absolutely convinced once this is coming, it has significant applications in the industrial metaverse. Because, think about it, you could do an overlay on your glasses with regards to inventory or assembly instructions. Or think in the medical field, of where to cut and slice, and what you have to do.
There’s a lot of very real-world applications that can help us make this much, much better. It will be an evolution, definitely becoming more realistic in that sense. AI is the wild card, headsets will be coming. There’s many more, edge computing, and such and such, but I will spare you all these details. But it’s there and we’re definitely convinced that there will be more digital worlds than what is there today.
Laurel: That’s a fantastic point to end on today. Thank you very much, Peter, for joining us today on this Business Lab.
Peter: Thank you so much, Laurel. It was a pleasure being with you today.
Laurel: That was Peter Koerte, chief technology and strategy officer at Siemens, who I spoke with from Cambridge, Massachusetts, the home of MIT and MIT Technology Review, overlooking the Charles River. That’s it for this episode of Business Lab. I’m your host, Laurel Ruma. I’m the global director of Insights, the custom publishing division of MIT Technology Review.
We were founded in 1899 at the Massachusetts Institute of Technology. You can also find us in print, on the web, and at events each year around the world. For more information about us and the show, please check out our website at technologyreview.com.
This show is available wherever you get your podcasts. If you enjoyed this episode, we hope you’ll take a moment to rate and review us. Business Lab is a production of MIT Technology Review. This episode was produced by Giro Studios. Thanks for listening.
This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff.
Humans and technology
The worst technology failures of 2023
The Titan submersible, lab-grown chicken, and GM’s wayward Cruise robotaxis made our annual list of the worst in tech.
Unlocking the power of sustainability
A comprehensive sustainability effort embraces technology, shifting from risk reduction to innovation opportunity.
Drive innovation with a tech culture of ‘connect, learn, and apply’
Offer an ongoing curriculum that aligns to strategic priorities and the latest technology to drive innovation, productivity, and social-good efforts.
People are worried that AI will take everyone’s jobs. We’ve been here before.
In a 1938 article, MIT’s president argued that technical progress didn’t mean fewer jobs. He’s still right.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.