Data, the Digital Twin and AI in the future of automotive – Transcript
In this episode of The Industry Forward Podcast, Royston Jones and Ryan Martin explore how data and the Digital Twin provide the foundation for smarter AI and software across the automotive industry.
Kate Eby: Hello, and welcome to The Industry Forward Podcast, where we explore key trends, transformative technologies, and real-world innovations that are reshaping fields from aerospace, industrial machinery, and semiconductors to pharmaceuticals and beyond. I’m Kate Eby and I’ll be your host for today’s episode. Today we’re diving back into the automotive industry and revisiting the world of software-defined vehicles. We’ll explore how digital transformation and the comprehensive Digital Twin are shaping the development of these vehicles and discuss how Digital Twin technology is laying the groundwork for artificial intelligence across the industry.
Kate Eby: Joining us again are Ryan Martin, Senior Research Director at ABI Research, and Royston Jones, Global Head of Automotive and Transportation here at Siemens. Welcome back to you both. Let’s start this episode with a question for Ryan. How important is the Digital Twin in the development of software to find and electrified vehicles?
Ryan Martin: It’s critical. The comprehensive Digital Twin is critical specifically. Reason being due to the rate of change and the complexity inherent in bringing these new products to market. It’s not just the OEM or the tier one supplier or anyone else. It’s really the change from these or statements to the and statements that makes the comprehensive Digital Twin significant because it’s design and engineering, sales and marketing, the OEM and their tier 1, Tier 2, and Tier 3 suppliers. And it’s getting all of these parties on the same page, the same timeline, and working in lockstep for the same shared goal through a common data structure, which really underpins the comprehensive Digital Twin.
Ryan Martin: And it’s these dimensions that really do create a significant competitive advantage. And it’s also what can help cut time to market in half and also ensure new products are more tightly aligned with customer preferences and regulatory changes.
Royston Jones: Yeah, I mean, I think for me, I think particularly when you look at EVs, I think the point that I was trying to make is that what you’re trying to create in the comprehensive Digital Twin is a twin that is a single twin that is not dependent on the type of performance attribute you’re trying to capture. So, you know, it basically can tell you when it’s too hot or when it’s receiving too much pressure or when there’s an electric current that’s much too high. I mean, a good example is that battery pack. It’s a multi-physics component, which is, you know, you need to capture that in a single twin, ideally.
Royston Jones: And the other thing that I was trying to emphasize is that Obviously, every one of these twins is going to be unique. It’s going to go through an industrialization process where it’s going to be stamped. If it’s metal, it’s going to be assembly. Then assemblies will not always be totally identical. So the industrialization process injects some imperfection and then When it’s out in the field, the driving styles, for instance, will be totally different. And what’s allowing you and facilitating eventually to create this comprehensive Digital Twin will be the data, because the data will give you an idea of this statistical variation that again, you’ll be able to build into the twin and provide a unique insight into the twin.
Ryan Martin: An important thing to emphasize here too is that there’s a ton of innovation happening, especially around Digital Twin and digital transformation. The reality of our situation is that innovation is not evenly distributed. And that especially applies to the world of data. We see 40 percent of companies still don’t have the capability to collect and analyze data in near real time and 70 percent cannot apply that data for prescriptive analytics, which is significant if you consider the benefits of having a comprehensive Digital Twin.
Ryan Martin: So data really underpins the foundational step for where industry is headed right now, which also really what I’m talking about is AI. So if companies want to implement and utilize AI without having a solid data foundation and a comprehensive Digital Twin upon which AI applications and contextual data can stand, it’s really not going to lead anywhere productive. So strong data foundation and having a more comprehensive Digital Twin are really critical for moving forward and trying to sort of lift all boats as we look for innovation to be more evenly distributed than it is right now.
Kate Eby: So essentially the comprehensive Digital Twin is producing a ton of data in and of itself. So how do you contextualize that or does the comprehensive Digital Twin provide the context to AI so that it can provide you with those insights?
Ryan Martin: Data and a contextual Digital Twin is the foundation. So the idea is that Digital Twins, there’s different levels of fidelity for a Digital Twin, from basic to intelligent and comprehensive. In a lot of cases, so I look at Digital Twins and even factory automation in a maturity scale that in some ways mirrors what we see for autonomous driving. So for autonomous vehicles, we measure them through levels one and five, 5 being a fully autonomous vehicle. In the case of a factory, we could think of level 5 being a fully autonomous or a lights out factory.
Ryan Martin: The reality of today is that most companies are somewhere between levels two and three. They started to embrace digital transformation, but they haven’t fully scaled it to all areas. And also, it’s important to point out in the case of a factory, there’s not a lot of scenarios where you need to have a fully lights out factory. Maybe semiconductor fabs is one of the few examples. Really it’s in isolated areas where you might consider higher degrees of quality and automation and safety as a driver to enable sort of a lights out work cell or area.
Ryan Martin: When we think about the comprehensive Digital Twin in this context, it’s actually imperative to have this robust full picture of physical objects, assets in process, not just in isolation, but in their environment to enable not only lights out work cells, but importantly, parts of the factory that are less fully automated, so to think about how everything works in concert with each other. And those are the kinds of capabilities in that symphony of physical and digital that is ultimately going to enable more purposeful and intentional AI scenarios from agentic to physical AI, which right now, a lot of the physical AI discussion is around robotics.
Ryan Martin: In the next few years, we’ll see that concept be stratified more to industrial automation as a whole. So spanning beyond robotics, and it’s really about applying AI at the edge, which a comprehensive Digital Twin is ultimately going to enable.
Royston Jones: So the way that I see it is that obviously within the comprehensive Digital Twin, there will be a brain and that brain will be the AI piece. So in a way, the data is not only making sure that the twin itself is going to be identical, but also the data will make the twin more intelligent. So a good example of that, again, is an EV battery pack where you have a battery management system, which is the software that controls the battery pack and the delivery of the energy. Well, within the AI, you can then start to make that software even more intelligent and more predictive as you’re capturing loads and loads of data for that in-field usage. So, you know, all these things fit together really well between the data, the comprehensive Digital Twin and the evolution of AI.
Ryan Martin: And there’s two aspects to the AI equation too. What we talk about, what we’ve been talking about so far a lot too is how data and comprehensive Digital Twins underpin AI use cases from copilots to generative and generative to agentic and physical. There’s also the role for AI to enable better data coherence between and across platforms and applications. So that’s something that’s actually a really important change when we think about Digital Twins evolving from basic to comprehensive because it now is much easier actually to implement a lot of these concepts that sound pretty space age and foreign. And AI just sort of lowers the bar to entry and also accelerates time to impact or time to value from the moment you decide to get moving with some of these newer capabilities.
Kate Eby: So everything we’ve been talking about around the comprehensive Digital Twin, about AI and the different applications of AI, that doesn’t just impact the car currently coming off the assembly line or that’s currently in service, right? Are those updates being fed back in of how the car is actually being used or how it’s performing to improve design processes for future iterations?
Royston Jones: Yeah, I mean, absolutely. I mean, traditionally, the design of these vehicles has been designed using idealized load cases. Now, what you’re getting is real world data back from over the air, which can help refine these various assumptions on, say, loads applied to the vehicle. So, that’s one area where you’re seeing some of that impact. I think a more simple example is windscreen wiper usage. If you’re basically looking at automatic windscreen wiper and all the options that you have to come on being automatic, getting that data and being able then to refine how that particular functionality is being used can actually lead to a better operation.
Royston Jones: Also, when you’re looking at driving styles and logging of the battery management system, you can then start to customize your own battery management system for a particular customer’s driving style. So lots of ways that this data can be used live in order to refine this product. And that trend will only continue.
Ryan Martin: It’s a great scenario too where real-world data can actually improve the user experience as well. So think about certain functions that are available to the driver of a vehicle and maybe some real-time road information gets fed back and the product engineering teams are able to push out an update that makes perhaps the driver safety sensors more responsive or more proactive and predictive. Perhaps the self-driving or assisted driving features, there’s some new software update that can be pushed out to the users based on new information that’s been collected in the field.
Ryan Martin: We see this in our smartphones all the time when you get a new app update or the latest operating system makes for a smoother interface. Those same concepts are applying to vehicles today, but on a much more on a similar scale, which is global, but with higher consequences and impact. And so these features are not just the ability to improve look, feel, and comfort, but also safety through real-world feedback loops. And then importantly, too, to ensure new products are more on target with what customers are actually benefiting from.
Royston Jones: A good feature in the UK would be for them to detect potholes that are about to come up. So maybe that’s an idea for some OEMs that are out there to get some sense of technology to avoid potholes, particularly in the UK?
Kate Eby: I don’t think that’s a problem just in the UK.
Ryan Martin: And the sample size is so much greater too. I mean, think about even a tire manufacturer, they do tons of testing on the rubber and they know really well through simulations and scenario analyses what the likely tread wear and life is going to be on those rubber compounds. But when you scale out the sample size to the entirety of the user base, you have much more data to work with. It may be, Royston, to your point, you can localize or regionalize then your products based on the way that customers are using them in their locale.
Kate Eby: I think that’s a great point to end on for today’s episode. Thank you both for another engaging conversation on all the exciting changes happening in the automotive industry. And of course, a big thank you to everyone tuning in. We hope you enjoyed the discussion and maybe learned something new. Be sure to join us again as we continue to explore the future of today’s industries. I’m Kate Eby, and we’ll see you next time on The Industry Forward Podcast.
Siemens Digital Industries Software helps organizations of all sizes digitally transform using software, hardware and services from the Siemens Xcelerator business platform. Siemens’ software and the comprehensive digital twin enable companies to optimize their design, engineering and manufacturing processes to turn today’s ideas into the sustainable products of the future. From chips to entire systems, from product to process, across all industries. Siemens Digital Industries Software – Accelerating transformation.
