Mercedes-Benz Group AG (MBG) Earnings Call Transcript & Summary

Good morning, everybody, and welcome to our Mercedes-Benz tech center here in Sunnyvale, California. And also a warm welcome to all of you that are joining us virtually from around the world. Today, we are going to talk about the Mercedes-Benz operating system, MBOS and lift the veil and look behind what is behind these 4 letters, MBOS. But it really starts with our strategy. We have set as the mission in our strategy. We will build the world's most desirable cars. If you think about the Mercedes brand promise for a second, what is Mercedes all about? It's always been this perfect blend symbiosis between innovation and technology on the one hand, but also modern luxury on the other hand. We want to use MBOS to fulfill this brand promise. Take it to the next level as far as digital experience is concerned in our products, but also in our whole company. So when I say we will build, how does that fit into MBOS? Well, we're a car company. We make things both in the physical world but also increasingly in the digital world. So if you're a builder and you start on this endeavor to develop an operating system from the ground up, you got to choose your role. And the role that we have chosen for ourselves is that we are the architects. We are the architects. What does that mean, to be the architect of MBOS? Well, it means we specify, we design, of course, we develop what this whole operating system is going to be all about. But we also procure and partner. If you're an architect, partial builder, you don't have to lay every tile yourself. You select the best partners around the world, and you work with the best to build this software house, this operating house. We integrate it into the vehicle, the sophisticated high-tech cars that we build. And of course, we keep it fresh with constant upgrades. Let's talk for a few minutes about what are the principles that we set out for ourselves when we made the first architect drawing of MBOS. Let's start with the first principle. Yes, it is purpose built, yes, it is Mercedes that controls the digital marketplace in the car. It is we that have the one-to-one relationship with the customer. But as I said, we work with the best. I see we have some of our old friends here in the audience today with Jensen and Sarah from NVIDIA and Austin, and everybody is here. We're going to have some new guests as well. But we're also going to announce a new partnership that is new to the team on MBOS, and that is Google. And Markus Schafer, who runs our R&D department, he's going to talk more about that together with Thomas Kurian later in the presentation. So first principle, purpose-built, but open to partners. It's really important. Second principle, and I talked about the one-to-one relationship with the customer. It all centers around a unique customer ID. So we are -- the product is directly in contact with the person. It gets personalized services. The car knows you. You know who you are. I mean, already today, if you take the Mercedes me app, I can open up here and I see my car, I'm driving an EQS 580. The car knows who I am. I'm -- my profile is stored in the cloud. When I step into my car, it recognizes me puts the seat in the right position, it chooses the radio station that I'd like to listen to. It puts the heating into the right spot, all of those different things. So this digital universe really centers around the customer and to be able to serve the customer, you have to have a relationship with a customer through a unique Mercedes ID, not just for the car, also for the enterprise. We will hear Britta on the marketing and sales side, talk about what this means in the wider context of the Mercedes-Benz universe. Third principle, privacy by design from the very beginning. We take data privacy, data security very seriously. That is why we are transparent vis-a-vis our customers with regard to this. In the Mercedes-Benz privacy center, you can choose your settings and decide what you want to share. And we have a, I would say, a basic philosophy always work in the interest of the customer. So when you receive data, what are you going to do with that data? You use that data to improve the experience for the customer, to improve the lives of the customers. So privacy by design is very, very important. And I think it goes without saying that in terms of cybersecurity encryption and so on, we use the highest standards that are available state-of-the-art in the market today. Fourth and last principle. If you buy a Mercedes, sometimes it could be like, I don't know, buying a very good French wine, it ages with grace, and it gets better over the years. But imagine now in the software space that you know the master winemaker, and for every year's new harvest, he comes with a couple of new bottles and puts it into your seller. That is what OTA, over-the-air upgradability, is all about. The product doesn't get old. It actually gets better over time. And we will talk extensively about this, especially when we talk about automated driving later on how you can decouple hardware from software and make sure that the product is alive over the whole life span of the car. So if I then put that together and say, let's try to sum up what is this architect strong? What is this building. It is what we call a service-oriented chip to cloud architecture. It sounds complicated, but I have 2 very qualified people, Markus Schafer and Magnus Ostberg, who is here today, our Chief Strategy Officer. They will go into all the nuts and bolts of what that means, and how we have done this from a technical point of view. For the customers, the domains and these are the application domains, that's usually what you see when you use the product, you have infotainment, the obvious thing is what most people think about when they think about the digital car. But it's only 1 of 4 domains, of course, automated driving, maybe the most exciting domain where there's so much going on, and we want to take it to the next level. Body and comfort, it is a physical product. We are physical beings. How do you create a well-being for the humans in the car. And then the obvious thing for a car, driving, charging, the battery management software, everything that's going on there. And it is a proprietary operating system, but it's open. It's open to partner with the best. Now have we done in the U.S. -- are we doing technology here for the sake of technology? No, we're not. It's all about the customer and to deliver a superior product experience for the customer. And if we do that successfully, we build economic value for the company and ultimately, our shareholders. So if you take that approach, you put the customer at the center of all things, you're also going to have economical benefits from it. Let me just give you a few glimpses of what I'm talking about. And then the other guys are going to go into more details later. Well, at Mercedes, if we now peak into this building here, what's inside the house. For Mercedes and for a car company, of course, the hardware is crucially important, desirable hardware. Think about the interior of your cars, maybe as your living room. You have now just gone out and bought the most expensive TV set that is in the market. But that TV set is not just sitting there alone, in a naked living room, you put the right designer furniture to go with it. When we launched now a couple of years ago, 1.5 years ago, what we call the Hyperscreen in the EQS and in the EQ and EQS family. Many customers told us that's insane. What are you doing here? It just blew them away, and it opens up so much more that you can do in the front, and we thought to ourselves, there's a hint there. Why don't we create a full dashboard screen pillar-to-pillar in the future for every single Mercedes so that every single Mercedes can have the option of going pillar to pillar, said and almost done. So starting with the first generation car that has MBOS and already in the sneak that we're going to see later, you will get access to hyper screens. I see some people nodding here that provide content to that and they're smiling. And I think that's the point. You can buy the most expensive TV set. But if there's nothing going on the TV just looking at a black screen, it kind of defeats the purpose of the whole exercise. So yes, we will put a world of entertainment into this. And it will be also specific for the different regions. I mean, if you live in if you live in China, maybe you're looking at different soap opera and if you live in my home country, Sweden or in the United States. So you got to go around the world and provide the infotainment content that the people want, and we're making some good programs rest on that. But because we're a car, we also need to talk about driving. How do you marry automated driving with navigation? Can you even get into a fusion here and make something that is even safe or even more intuitive, even better, even smarter. We think you can. So I think this canvas, it can be so many things. It is a window into a world of opportunity and experience for our customers. And yes, for the people in the room, one of the one more thing moments of today, we're going to give you a sneak preview of the new E-Class. So you can look at the interior of the new E-Class and the big screen and everything what's going on here. That in itself is big news, but even bigger news that some people might not know is that actually this is the precursor of MBOS in the infotainment domain. Our team -- our R&D team has programmed the complete software for this infotainment domain. For the first time company's history to take on that big of a software task and do it comprehensively. So yes, Magnus, Markus, the pressure is on you guys now to kind of do the finishing touches here. We're going to launch the E-Class later this year. And Markus is going to give us some glimpses of what's behind that. What are we going to offer? As I said, maybe even more exciting the autonomous drive domain. We have been working on assisted driving and assistant driving system for more than 20 years, almost 25 years. We've always been looking at this, how can we make driving safer? How can we make it more convenient? How can we put all these little angels in the background? If you get into a situation, the car helps you negotiate the situation. And in a conversation that I had with you, Jensen, some -- a few years ago, we sat and talked about the philosophy of where are we going to take this next. And here comes a big change to the technological strategy and the business model that we have had previously. From MBOS 1.0 forward, we're literally going to put super computer-like performance in every single Mercedes. And for those of you who don't know the MMA architecture, it's an internal code that we use. It's actually the next-generation entry luxury vehicles in the Mercedes portfolio. So it's like where you get into the Mercedes brand and then you graduate up. And eventually, maybe you end up with an S class or something like that. So every single Mercedes from that point forward will have a supercomputer in it, number one. Number two, every single Mercedes will have a very comprehensive set of sensors, the eyes and ears of the car to figure out what's going on to be able to deliver assisted driving and automated driving. And that sensor set, the base sensor set in future cars, even in the entry vehicle will be significantly more comprehensive than the most expensive sensor set that you can buy in the S class today, bar maybe the Level 3 vehicle, which is in a little bit of a special category. And the third thing of the business model is when you have separated hardware from software, over-the-air upgradability, the product never gets old. So we can just keep on developing, developing, developing and make sure that the customer gets a better product. And this is something that we're doing with NVIDIA. Let's just have a quick look at this video here. [Presentation]

I would like to invite our business partner good friend, Jensen, Founder and CEO of NVIDIA to the States, Jensen. Good to see you. How was your trip over here this morning?

Well, first of all, I came in an EQS, my own EQS. I have 2 EQSs. The first 2 in the United States, I understand, one for my dad and one for myself. And if I could just tell once short story real quickly. When NVIDIA went public, the thing I wanted to get for my parents more than anything was a Mercedes. And I bought them a Mercedes S500, when NVIDIA's stock price was $13. And my mom to this day tells me it is the single most expensive car any human has ever purchased. And so you should...

So you could have...

I should have given you stock. You should have received stock instead of money. And anyway, it's a $200 million Mercedes. My father still has it.

Maybe even more expensive than the French wine. Well, Jensen, when we sat down and talked about this and we decided, let's go into a strategic partnership. What was your vision for us working together in this space and taking automated driving to the next level. How do you see this playing out?

Well, first of all, you -- what you're doing with Mercedes is really incredible. This is a luxury company now being infused with software through and through. And the thing that is really amazing about the decision that you made to put AI supercomputers in every single car and to enrich it with a cocoon of safety sensors around it. so that every single car becomes part of the Mercedes fleet for the rest of time. You're building up a fleet, a giant fleet of upgradable cars that, as you said, never gets old. This endeavor is really quite amazing. The first thing is we start from building the most ambitious, safe computer the world's ever known. From the ground up, from the chip, to the system, to the operating system, MBOS, to all of the applications on top, even the artificial intelligence, AI, models are designed to be functionally safe. Safety principles from the bottom all the way to the top. Second, I just mentioned the cocoon of sensors. Third, this entire car is completely programmable, and we're building the car computer and the cloud computer together, chip to cloud so for the rest of time this fleet will be upgradable.

That is very exciting. When I visited you on Monday, you -- let me peak into some of the stuff that you guys are working beyond this relationship here with autonomous drive. And we have a couple of pilots going now for Omniverse, building the virtual factory before we even build a factory, tell us a little bit about Omniverse. Well, what's that all about? And how can this help us and others leverage their business?

Well, you're reinventing Mercedes, the automotive industry from the beginning to the delivery of the car to the support and the updates of the car from end to end, from styling to design to engineering, manufacturing, the digital smart factories, the robots in the factories that build the cars that are robotic that will be updated with software. Your entire company is going to be software-defined from end-to-end and artificial intelligence powered from end to end. And this is one of the thing that's really, really quite amazing about the digitalization strategy that Ola and the management team at Mercedes has for Mercedes is going to reinvent the company to become a luxury-focused technology-driven, software-driven company. And I think it's just really a great pleasure to be part of it. And for us, the way we participate is creating a system called Omniverse that understands physical things that, of course, Mercedes builds, and the software world and the artificial intelligence world that we are in, Omniverse allows us to bridge that, to fuse those 2 worlds from end to end, from beginning to the end.

Again Jensen, good to see you.

All right. Have a great day.

Thank you, guys. On that note, I was talking about MBOS being all about delivering a superior customer experience. But as this one example here with the Omniverse illustrates, it actually is more than that. It is a platform that connects all parts of our business. So if we go from the architect's drawing of this one house and look a little bit at what the neighborhood is looking like. In a simplified way, you could say that this is your neighborhood. You have MBOS at the center of things. As we said, it's a chip to cloud architecture. So our so-called Mercedes Intelligent Cloud is what holds the profiles of the customer and so much information and intelligence sits there and is available around the clock and then when we have upgrade features over-the-air. And of course, it is the device you can control your car via the device or you can hook up your car to your smart home and do all sorts of different things with different devices. But it also has connection into the enterprise. So if we talk about product creation, operations and commerce. The whole point of using Omniverse to create a digital twin for manufacturing is before you pour the first concrete of slab, you have actually built the whole factory digitally and you have tested every process. When you look at it on screen, if you would take a photo of the factory and you look at the digital representation of it and you take away the label, which is which, you really have trouble to distinguish, which is which. That idea is something that we have been working on, Markus, in engineering for many, many, many years, moving the boundary of what simulation can do and go into hardware later, then we can be faster in development cycles, and we can also reduce cost in the product creation process. And Britta will touch upon commerce, omnichannel, how does this link? I mean it is obvious right that if you have this unique Mercedes ID, you don't just use it for the car and for the app, you use it in your CRM systems and in every touch point in the Mercedes universe, if you can. But enough for me now. I know you're excited about seeing the details of the technology. Markus, Magnus, the stage is yours.

So thank you, Ola, and welcome, ladies and gentlemen. So our house is built to last, which means, it must be fit for purpose, not just for now, but also for the future. And that is the vision that we had when we set up out to develop MBOS. We knew that we achieved something like this scale and depth, the only way is to be in control of all levels, to take a completely new approach. And as Ola said, to be architects of our own house. Magnus and I are going to talk about the details. What means to set up MBOS. We will explain exactly what MBOS is and show you how it will help us to take our customer experience into a new dimension. So let's start with the what. Technically, MBOS is a chip to cloud architecture. That's because we see software and hardware as 2 sides of the same coin. At the same time, as the architects of MBOS we are able to decouple them. This gives us great flexibility actually in 2 aspects. The first relates to time we are able to separate the software and hardware innovation cycles. And the second relates to scope. We can use one platform across our entire portfolio from small cars to large vehicles. In our house, MBOS as a technical foundation on the one hand and the luxury living space, Ola, you explained that on the other. So let's start with the technical foundations and what they are made up of. So our chip-to-cloud architecture, our scalable proprietary platform, it enables us to decouple hardware and software and is designed to ensure safety and privacy, Ola you said that. And to talk about this in detail I would like to introduce my dear colleague Magnus, the Chief Software Officer. And Magnus, you joined us in September 2021. And Magnus brings nearly 25 years of know-how in automotive software from infotainment, telematics, satellite architecture, automatic driving and I think way more than that. And I think you have already made your marks designing and being the architects of MBOS. So Magnus over to you and to the foundations.

All right. So in this is an entirely new approach for us here at Mercedes-Benz. It will speed up software and hardware development and allow constant flow of innovations into our vehicles. And that means a better product for our customers. So starting with our MMA platform, we will have our own standardized software and hardware-based layers. So let's start with the hardware base layer, which is the sign on a blank sheet of paper. So its architecture is service-oriented, and it contains the domain computers and the connectivity module and they host all of the intelligent functions in the vehicle. The specification of each domain computer is precisely tailored to its functions and performance requirements where the services are abstracted to the domain computers. The GPU, our infotainment computer, can power more than 3 high-end gaming monitors. And the hardware architecture of our automated driving computer is capable of ASIL-D. So that failure rate is close to the flight control system of a commercial airplane. So besides these high-end computers, our architecture design includes the controllers for exterior and interior driving and charging functions as well as the connectivity module to the Mercedes-Benz Intelligent Cloud. So with this architecture, it gives us a deep integration of the vehicles and the central control of every sensor and actuator. So due to the standardization, it's scalable, right? It scales from compact cars to full-size cars like Markus just said. And at the same time, it decouples the innovation cycles of the MBOS from the vehicle platforms. And of course, our hardware approach goes hand-in-hand with semiconductors. So as the architects of the MBOS, we are deeply involved in all aspects of the ship strategy. It's completely different to the historical model of any OEM. So our infotainment and automated driving computers, they do have state-of-the-art high-performance SoC like Jensen and Ola just said. But also in the microcontroller space, we are maximizing standardizations while driving performance and efficiency. So we are able to cover more than 90% of our EE architecture with chip families from a very, very small number of suppliers. So with that, we have categorized all the chips and derived tailored strategies. So these strategies consider a few educations of the requirements as well as the learnings from the ship crisis. Key elements are direct sourcing, long-term sourcing and increased in-house development. So in the high-end computing cluster, we are driving even further engagement into the chiplets technology. So we see the potential of chiplets primarily in the energy efficiency, latency, bandwidth and flexibility. And to maintain economics of scale, we have recently became the first OEM to join the UCIE consortium. That is in order to push standardization for chiplet interconnects. This will be instrumental in setting industry standards for automotive for us, in particularly those used in automated driving technology. So that was all about the base layer. Let's then take a look at the software base layer. Here, MBOS also marks a radical step change. So we are leading its design and construction, maintaining overall control but using proven standards. So between the domain computers and the connectivity module, we use an automotive Ethernet connection. So this is a high performance up to 10 gigabits per second. It enables full updatability of the entire MBOS platform within minutes. So communication is based on vehicle APIs that are independent from our car lines, as I said, and it contributes to the scalability of MBOS. So together, the building blocks of our software-based layer enables service-oriented communication for our applications, diagnostics, other updates, security functions and across the entire MBOS platform. So let me now guide you through the key principles of our software-based layer. And I'm going to use our infotainment computer as an example here. So starting at the bottom, the first sublayer contains the driver and security functions. These enable Ethernet communications as well as firmware and hypervisor that sits on top of the SoCs. On top of that, we're starting the operating system components. Here, we're using open source Linux and it gives us the development speed for new functions with a very robust rollout and capability options. On QNX, we run the safety critical functions for the driver displays. And then we have Autosar Classics, which we jointly adapt with our partner, Vector. So on top of that, we have the middleware layer. It's designed to enable flexible integration of features developed in-house as well as third-party content, including regional heroes. To integrate Apps then seamlessly on top, we are using Android container interfaces. It is especially relevant in markets where Android apps are increasingly popular, such as China. So I'm going to zoom out now for a moment and look at another building. This is the Mercedes-Benz Intelligent Cloud that Ola talked about. It is really the game changer with this new approach. The Mercedes-Benz, Intelligent Cloud is a modular connected service platform that opens up a world of new possibilities. And it's also the backbone of our software tool chain, right, along the entire vehicle life side were using this tool chain. So let's look at our cloud infrastructure, starting from the bottom. Today, we host our worldwide cloud solutions from hyperscalers such as Microsoft Azure and Amazon AWS. In the future, our multi-cloud strategy will be complemented by partnership with Google GCP for the West and Tencent for China. So on top of these hyperscalers, we build a container-based micro services mesh platform using Zero Trust concepts. It is made up of 4 layers, each which is all functions, which I'm not going to go into detail now. So -- but what we do though, we develop more than 80% of the software for the Mercedes-Benz and [ Selte ]cloud stack in-house. This gives us shorter lease times for new features and an endless learning opportunity based on customers' usage. So owning the Mercedes-Benz Intelligent Cloud gives us the data, right? And this data gives us many, many benefits. One of them is the ability to perform targeted data collections and analytics to improve our products and services. It is how we leverage this data that it's key to digital luxury. So by 2025, we will have access to data for more than 16 million cars, up from the 10 million that we're already hosting in the Mercedes Intelligent Cloud today. And because we own the system, we also have full transparency over the use of customers' relevant data. So with the Mercedes-Benz privacy center, the customer has full control of all the data, be it personalized or anonymized -- so -- that is how we leverage the data. Now where and how are we doing all of this? Well, at the center of our development network is our Electronic Software Hub in Sindelfingen. This is where we do all the integration, validation and testing. So the Electronic Software Hub was built to facilitate workflow and communications, super close to R&D as well as our headquarter production experts. So from top to bottom, the process starts with the integration of software code, right? So the next level down, we interact directly in a virtual world before the physical hardware components are added. The lower floors then is where we then carry out complete mockups and vehicle testing. So the Electronics Software Hub is a highly efficient integration machine. It brings cost savings as well as help us leverage our global team. So speaking of Global, we are maximizing our software and hardware competence within our networks of hubs all over the world. Here in California, we emphasize on autonomous driving software as well as UI-UX design and telematics. Tel Aviv, we have expertise in cybersecurity. And in Asia, we really focus on our regional heroes and local content. So that rounds up the foundation of the MBOS house. So as the architects of the MBOS, we control our service-oriented chip to cloud architecture are scalable, it's decoupled. It gives hardware, the flexibility to live its separate life from software. So MBOS is a proprietary platform built on standards. At the same time, we retain the sovereignty of all aspects of strategic importance. This includes management of our own data tree, as I said, as well as safety and privacy. And finally, we have the freedom to work with our partners of choice. So I'm going to hand back to Markus soon to explain what this means through the living space. But first to wake you up after my long speech here, let me give you a taste of what these great things have in store with this video. Enjoy. [Presentation]

So all of this will make our house the most desirable in town because the living space is where our customers can really feel the benefits of our vehicle functions across our 4 domains. So let me start with body and comfort, where we are creating a truly personalized luxury experience. To do that, we need control of all actuators and sensors every single one MBOS gives us that. To assure you, this would not be possible with a third-party provider. With MBOS, we are able to create a true multisensory experience by orchestrating all vehicle functions from video, motion, light, climate control and even fragrance. The result can be a stunning immersive entertainment experience. And those of you here with us today in Sunnyvale will have the chance to find out what that feels like in our demo vehicle outside. And I promise you, it will blow your mind. Another area where this overall control via MBOS can benefit our customers, is our energizing comfort and energizing coach. We are taking them to the next level by enabling connection to wearables. Our energizing app uses data such as heart rate, stress level and sleep quality to recommend a personalized comfort program. Likewise we can bring together comfort functions such as massage, active ambient lighting and 4D surround sound for enhanced well-being. And if it is not enough, for added convenience, we can also offer a digital vehicle key for smartphones. Coming now to the broader domain of infotainment, and when it comes to automotive, digital luxury, first-class route guidance remains absolutely a top priority. We have been driving innovation here for many years, providing our customers with the latest solutions. Today, we are announcing a major step forward. For almost everyone in North America and Europe, Google Maps is actually the go-to. So now we are bringing it into Mercedes vehicles. And our collaboration with Google in geo-positioning takes in-car navigation to a whole new level. It's a world-first. We are reinventing route guidance experience by bringing together the very best from automotive and tech sectors. So to talk more about this, I would like to welcome Thomas Kurian, the CEO of Google Cloud, Thomas.

Thank you.

Tom, it's great to have you today with us here on stage. Actually, we are neighbors, right, for many, many years. And we did a couple of projects over the last couple of years, but now we're taking it really to the next level.

So thank you for having me. It's a privilege to be here and delighted with the collaboration we've had.

So you bought your team with you and we're working with the team now for many, many months out day and night actually working on new products, and it was the target to create a new customer experience. So we're talking about the new products, and let me start with the first one, place details by Google. So I would say POIs,but it's place details by Google. So can you tell us what this would offer to our customers.

It represents every Mercedes car will now have the trusted, reliable information about 200 million places around the world. So if you're a driver, you can, one, find the place that you're looking for. If you are not totally comfortable that, that's the place you want to be, you can look up a rating and review if you drive up and you're a little bit trying to find it, you can see a photograph of the place, and it brings together the trusted information that so many people are comfortable with Google Maps with the luxury experience that they have in Mercedes vehicle. And most importantly, we're doing it right now, isn't it?

Yes. Just look at this video here, and that's the beauty of it. You don't have to wait. So our custom system is able just to incorporate this feature as we speak, exactly. So -- but that's not all. I mean we're expanding our partnership. And the next step for us in the next years is to go to Google navigation. And that's going to open up a whole new world of experience to our customers. And as we said, that's the trusted source in the Western world for our customers, and that's the move we are going to make. So what are the advantages there?

Well, you're our first -- very first partner in this initiative. What we're doing is bringing the next phase of evolution of Google Maps platform with the amazing display and navigation experience in Mercedes vehicles. So think about, somebody's driving and they want to find the most optimal route. They will be able to do that. If traffic conditions change, the car will automatically adjust and give them a better route to take. -- if they want to take information, for example, turn-by-turn instructions, they can do it with voice-assisted steering. If they're approaching around about and they want to get guidance on how best to navigate the roundabout the system will automatically help them by adjusting vehicular speed and things like that. So it's really taking the notion of place that we're starting with to now include all the information that a system like Google Maps can provide, but integrating it into the way that assisted driving can help people and with the amazing display, providing very high definition visualization as well. So we're super excited to have you as our very first global partner in this, and we think we can redefine how people experience the vehicle and also driving itself.

So as I said, it's a brand-new product and joining forces there to give luxury experience in navigation, but keeping the beautiful canvas that we offer in the car. So combining best of tools actually here. So then another way of collaboration is a new home field actually on Google Cloud and using data and AI to enhance a driving and navigation in the future. So that's the next -- what we explore together. So can you tell us what we are doing there.

Well, a couple of things. One additional thing we're doing with you is bringing YouTube into your infotainment experience so that people can watch and experience all the amazing stuff that is uploaded every second of every day on YouTube, they experience it in Mercedes vehicles, something we're really thrilled about. More and more, as I listened to you talking about the vehicle, it's becoming more and more software powered as well. And there's a lot of data, fleet data coming off the vehicles. We're also thrilled to be bringing Google's AI and cloud technology to help you, one, build AI models faster. They can help you understand customers, driving experiences, et cetera, but also give the data platform for analysis and to look at vehicular data in a new way. These are all the next stages of evolution as you software power vehicle, we want to be the partner giving you the platform to do all of this with. And so we're super excited. Our teams are thrilled to be working with your team. We think the combination of your strength in a premium luxury vehicle and our technology is a great promise in all these different areas, and we couldn't be more excited to be working with you.

So I couldn't explain that better. Thank you, Thomas. So I promised my team at home that we are virtually pressing a button now for our first product, which is places. And that's exactly what's going to happen. I mean they're flipping the switch now in Germany, and we're going to have this over-the-air download worldwide and can you imagine that gives me goose bumps. So they're downloading and all these little small Gs are dropping into the screens of our customers, and they will have a great, great service of Google Places now literally today. So great to have you here.

Thank you.

Looking forward for our cooperation...

Thanks a lot.

So ladies and gentlemen, as you have heard, this strategic partnership will definitely keep us at the forefront. So we will work with Google in those markets where they are dominant. That means North America, Europe, Australia, Japan, India and others. For other key markets, we are working with regional heroes, for example, AMAP in China and soon TMAP in Korea. So we are bringing the best real-time 3D graphics into our car and combined navigation with stunning experience. So one example is the highly advanced technology from the Unity game engine. In China, we aim to deliver an enriched UI for automated driving functions based on MAP data from AMAP. And this includes a fantastic lane-level map view and soon, it will be able to offer navigation-based system initiatives, lane change. So I'm going to hand over to Magnus now to explain in more detail what MBOS means for the other aspects of infotainment and experience. Magnus?

All right. Thank you, Markus. So we are keeping UI at the forefront. The MBUX Hyperscreen was just the beginning. So for instance, a pillar to pillar, one piece display, as shown in our vision, EQX outside will find its way into our series production vehicles. But UI is more than just displays. So our Hey Mercedes voice assistance is here a great example. Data shows that customers already interact with our Mercedes VA several times a day, and those figures are growing very, very fast. So MBOS enables us to anonymize data from those interactions paired with AI to keep evolving this to create a more human-like conversation. That's why we're introducing Just Talk. This is a feature coming up in our next MBUX. So when customers are alone in the car, they don't have to say, Hey, Mercedes, to activate the voice assistance. We think that is going to be an amazing new natural experience with the car. So we also think of these touch points in our living space here, a bit like UI in our vehicles. So -- there are many, many partnerships that come into play to create the services and the features and the content that our people love. So that's why we invited a list guests into our living space. So everything from video streaming, music, gaming, conferencing and social media. And to do this, we are building a system that makes it fast and efficient to enable fresh content and third-party apps. And that we are doing by using what we call the MBUX API for Android. This enables us to implement a large number of Android-based apps very efficiently. So for example, we have a dedicated app offering to meet the China preferences. And in the Western market, our customers can look forward to a steady stream of new apps fueled by the FORVIA Group with the Faurecia automotive app platform, which is an independent Android app ecosystem. Our vehicles are already equipped with the best-in-class displays, such as the MBUX Hyperscreen and rear-seat entertainment. So our customers will have their favorite video content on these beautiful canvas wherever they are in the world. In China, we will offer a regional specific content via Tencent Video. And because entertainment and digital technology are extremely important in China, we are using our Abacus AI engine, which is the scenario-based AI integration platform. It provides an experience that is just right for China. This includes video recommendations and hot playlist on QQ music. Across all markets, we are constantly improving our streaming offer and with exciting new content. We will also upgrade our hardware situation with Amazon's Fire TV stick for rear-seat passengers starting this year as an add-on. And as you know, we have already integrated video content provider ZYNC with more than 400 titles and nearly 2,000 hours of run time. So here's the short message from our ZYNC CEO, RJ. [Presentation]

So RJ, thank you for being here in person, and it's about content and speed of implementation by the way. So just like video streaming, MBOS will allow us to open a whole new world also of in-car gaming. We already offer mini games such as Tetris, Sudoku and payers. But with Antstream platform of arcade games, our customers will soon have larger access to 1,500 classic games and 500 mini games, and there's more to come. We want to bring AAA gaming into our cars -- and stay tuned. We have some news for you pretty soon. So speaking of Tune, our music collaboration with Dolby, Apple Music and Universal Music Group is already bringing unmatched audio excellence. And we intend to introduce this unique feature to our entry segment too. In the first for Mercedes, we are bringing video conferencing into our vehicles, integrating the popular services of Webex and Zoom. We intend to bring digital art into the car for the first time. Our mode for digital art will have a unique UI design tailored to the car as well as customer preferences and MB exhibition feature will enable customers to create their own private art gallery. They can choose the art collection created by Mercedes-Benz -- so what's more with my NFT feature, we enable them to exhibit their own NFTs in the car. A great living space is somewhere you spend time doing the things you want to do. We feel the best way to give that time to our customers is through automated driving. So our drive assistance systems have evolved over many years because we focus on safety. Our latest Level 2 development is the system initiated automatic lane change feature for an even better highway experience. It starts this year in North America, followed by China. Further enhanced Level 2 systems are in the pipeline. And we are at the forefront with our pioneering achievements in Level 3 and Level 4. We were the first manufacturer to put a Level 3 system with internationally valid certification into serious production. In Germany, our Level 3 DRIVE PILOT has been on the market since May of last year. This year, DRIVE PILOT received certification from the state of Nevada. And very soon, we expect California to certify it too. This generation of DRIVE PILOT enables conditionally automated driving at up to 60 kilometers per hour in Germany or 40 miles per hour in the U.S. were allowed. We are working to increasing this using a step-by-step approach with safety always as the first priority. Next year, we aim to increase speeds to approximately 90 kilometers per hour on German autobahn. That's roughly 55 miles per hour. And to achieve this, our sensing technology is based on multiple systems, including radar, cameras, LIDAR and more. redundancy of major components is an important factor here. And I should point out, we are not just working on automated driving for the road. We are also at the forefront of automated parking. Our intelligent park pilot is the world's first Level 4 parking system. That's the current status. So now we're going to look at the important role of MBOS in the future of automated driving of Mercedes-Benz -- with our in-house competence, we have made great strides in software and hardware developments, but to expand the use cases for conditionally automated driving, we have to go from a rule-based approach of today to a machine learning approach in the future. For this return to external partners with lots of experience in their field. A new generation of our Level 3 system is already at an advanced stage. It has also been optimized for quick rollout in terms of markets and models. Our strong partner here is NVIDIA. I will hand over to Magnus for some further details. Magnus?

All right. Thank you, Markus. So a core feature of MBOS is that it will enable lifetime learning powered by data, as I said previously. Our system will enable to learn on the road taking advantage of our best-in-class sensor setup across the entire vehicle fleet. Our vehicles are able to collect up to 300 petabytes of data every year to further train our deep neural networks. And our adaptability in vehicle AI automatically chooses a fraction of this data that is the most useful to improving our software stack. It means our next-generation automated driving will learn about and handle infrastructure changes and continue to improve throughout its life. So we are working on algorithms and software stack that will provide completely new driving and parking functions. This will be able to deal with the diversity of different geographics from Europe to China, U.S. and beyond, and they will continue to improve their performance, adapting to safe, human-like private behaviors. And as Markus said, we are working with a very strong partner to realize all of this. So to look at it in more detail, I'd like to welcome Sarah Tariq. Head of Autonomous Driving Software at NVIDIA.

Thank you. My pleasure being here.

Great to have you here, All right. So let's just get into talking about software architecture because that is our passion. Both of us are sharing. So can you please elaborate on the architecture that we're now building together?

Sure. So we're building a full stack EV platform that's everything from the OS to all of the middleware and the software on top that gives us the capability of doing advanced autonomy and all sorts of different LDDs. The key aim of our architecture is to create software back in comfortably, but also safely drive our customers in their day-to-day commute all over the world.

That's a great note there on safety. So safety is a big concern. It's always in the forefront of Mercedes. So what are we doing really then to ensure safety in this new approach of machine learning?

Right. So like Jensen said, all of our architecture from the hardware, the SoC, the OS, middleware software, everything is built, certified design to be safety first. In addition, the state-of-the-art algorithms, the AI that we're running on top, it's built and trained, not just to deal with the expected what should somebody do, but the unexpected, what could happen. In addition to this, we have a lot of monitors and redundancy algorithms that are checking inputs and outputs, making sure everything is going as expected. And again, like Jensen and Ola talked about before, we have the sensor suite that is redundant, right, like cameras, radar, LiDAR and 360 all around the vehicle, and this provides us with an extra level of security safety and the data that we are seeing from these sensors up to a terabyte an hour per vehicle, we can also use offline to further harden our system.

Right. Now speaking of data then, how exactly is the data this immense amount of data that we're collecting from the censors actually then being used?

Right. So data is the lifeblood of complex AI, right? The very first thing we do, all of these networks that we're running on our vehicle, everything from perception, understanding what's happening in the world to predicting what people are going to be doing to generating the automated maps that you're issuing. All of that requires training with all of this data that we're seeing worldwide. And then the next step, we take all of this data and we create from it offline tests and metrics, everything we can do to ensure that any time we're making an update to the software, we can test it off-line against these scenarios even before you put it on the road.

Right. And to doing this speed simulation is a very, very key factor, right? So give an example of what we're simulating.

Right. So I just talked about using all of the sensor data to replay and test whatever we saw in the world. But actually, like we were discussing before, we are using Omniverse to also generate all sorts of new scenarios a vast variety of scenarios that you could encounter in the world, everything from differences in weather conditions, locations, people's behavior, different scenarios, we can simulate all of that end-to-end synthetically. And we can also simulate the sorts of scenarios that you never want to see in real life, dangerous things people sitting on the road and so forth. We can simulate them and we can test against them. -- and all of that is a focus on our safety.

Excellent. So conclusion, the more the system learns the better it gets, right?

Exactly.

Okay. Excellent. So as you see, the partnership with Mercedes-Benz and NVIDIA is already living up to its promises and has potential going forward is just huge, right? So with that thank you very much, Sarah.

Thank you. My pleasure.

Yes. Thank you, Sarah, and Magnus for this insight actually. So starting with our new MMA platform, we are moving, as we said, from the rule-based approach to an AI-powered and data-driven approach. So together with our sensor setup, including redundancies, this enables us to meet the high demands of urban environments. And it will be augmented by a crowd-sourced HD map to handle the most complex situations. The result will be an enhanced Level 2 system for automated driving. So we intend to progress towards point-to-point assistant driving active lane changes based on navigation. And that includes the very start and the end of journeys where we plan to introduce new parking features. All of these calls for highly advanced technology and, of course, LiDAR technologies in combination with radar and cameras. So we are convinced that redundancy is the right approach for Level 3 automated driving and beyond. By using the best-in-class LiDAR tech from Luminar, we can achieve high range for the smallest and see smallest objects with a low reflectivity. Dynamic steering of high resolution also allows us to focus on specific areas similar to the way human vision actually works. The sensor setup for our Level 3 system will be even more comprehensive. We will have 2x the compute performance of our Level 2 system. There will be additional redundancy in sensing modalities, braking, steering and power supply. So this will help us to ensure our high safety standards and customer satisfaction. So customers will be able to benefit from the Level 3 system automated system at higher speeds. Our ultimate goal is 130 kilometers per hour, around 80 miles per hour. So functionality will include automatic lane change and highway to highway transfer -- we intend to roll out these features worldwide. So now I would like to invite Luminar CEO, Austin Russell, for some further discussions on this stage. Austin, please join me. good to have you. I like your shirt, Austin already. Also taking a short break from, I think, very, very intense day of developing LiDARs often we talked about the importance, and I talked about the importance of LiDAR for Level 3 driving. How do you see that?

Yes. I mean, LIDAR is definitely a critical part when it comes to autonomous driving more holistically. And the whole point is that this LiDAR sends out millions of pulses of this laser light out in the environment, knowing exactly how far away the objects are gives you a ground true view of everything going on. And that accuracy and reliability makes all of the difference. 99% actually gets you so far. But when it all comes down to that last 1% for autonomous driving and understanding the implications of what's around you, it makes all the difference to have that. And that's where I think it's great to be able to see Mercedes take a lead just as being at the pinnacle of performance and safety on the car side to be able to do that same strategy when it comes to autonomous driving and as well as assisted driving.

Well, but when we met, it was not only about getting a LiDAR, right? The best possible LiDAR, so we want to drive fast maximum speed there, highest safety. So that's how we came together, right? Looking at your technology. And your technology is called Iris and just explain what the Iris LiDAR is.

Yes, absolutely. So this Iris LiDAR, basically, you can see -- I saw yes, you have a blow out of it on here, too. You can see it out on screen even through the transceiver through every aspect of it. We're actually delivering this next generation of the Iris technology, specifically to Mercedes here starting out as a lead customer for this that takes it even a step further in terms of functionality for performance and safety and capabilities. And the key is being able to now operate for autonomous driving functionality not just for lower speeds, but also being able to move it to higher speeds, while at the same time, being able to substantially improve the performance and safety around assisted driving systems concurrently. So that's the beauty. It's the same hardware that can help enable all of that. And of course, when in conjunction with the right compute systems like the NVIDIA systems and the software that they're doing to enable that, it makes all the difference.

And I think it's important we work as a team here. So NVIDIA Luminar, Mercedes, this is real teamwork, right?

Absolutely.

And I mean the whole name of the game is to scale also. So not only one car, we want to equip of course, with LiDAR. I mean we intend to do it really in a broad range of vehicles. So scalability is important. And again, it's not your first generation of LiDAR. It's the next generation of LiDAR. And the idea was to make it really, really slim. -- a nice fit to the cars, not to destroy. Gordon is here, our Chief Designer on right, not to destroy the nice beauty and aesthetic of the vehicle. So we're making it really, really slim absolutely functional and the idea is to scale it, right?

Absolutely. And I think that's the whole point around what we're doing together. It's one thing to be able to put it on one flagship vehicle to be a little to start out. But I think the key is that you guys have this bigger picture vision and accelerating that across the, as you said, a broad range of vehicle lines. And that makes all the difference because when it comes to safety and autonomy and making it accessible for everyone and all the customers, that makes all the difference. So we're -- we built this LiDAR in this architecture with scalability and buying from the ground up, signed for manufacturability. We built this to scale starting off into millions of sensors and ultimately tens of millions of sensors with this next-generation LiDAR architecture. So that's what makes all the difference there to be able to scale alongside Mercedes here. And with that, we'll actually even be expanding our capacity and even building out a new facility in Asia to support the additional volume. So it's big.

So R&D is important. -- industrialization is important. That's all what we are carrying for, right, as a team here.

Absolutely.

Austin, Great to have you here on stage. Thanks for being here. Thank you.

Absolutely. Great to see you guys taking huge leadership.

So ladies and gentlemen, if you have some energy left we're going to the -- basically a last domain and the last domain is driving and charging. And that contains the system, all the activators relating to the drivetrain. It includes, of course, the battery management. The combination of route guidance and range management is also very, very crucial to the user experience. The range is function in our EPS is a good example of our approach emphasizing our own operating system actually allows us to access all vehicle data, for instance, state of charge, energy consumption and so on. This would not be possible with a third-party system. By aggregating all available data, including environmental information, maybe such as wind speeds, we can achieve a whole new level of range prediction. And by integrating the charging infrastructure into range management, we can create a seamless first-class charging experience. Our global Mercedes-Benz high-power charging network starting here this year in the U.S. is a perfect example. As we announced at CES in January, this will be one major step to globally supporting electrification. Our efforts were even acknowledged recently by the White House -- so MBOS will help ensure it's a fully connected service. So to sum it up, as the architects, we can report that our MBOS built project is making excellent progress. Our living space is already well ahead of schedule. MBOS gives us the freedom and scope to create a personalized luxury experience. With the right partnerships, we are providing our customers with the best content and navigation. So MBOS can help us to retain our leadership in automated driving and ensure intelligent and seamless driving and charging. As Martin said earlier, we have a strong foundation based on our service-oriented chip to cloud architecture. It is a highly scalable architecture and enables us to decouple software and hardware. MBOS is a proprietary platform built on standards designed to ensure safety and privacy. It also gives us the maximum flexibility to work with our partners of choice. MBOS is a game changer for the company and our products. But most importantly, for our customers. In a moment, my colleague, Britta, Britta Seeger, will talk about the customer perspective in more detail. But first, we have something very special for you. You see our latest generation of MBUX is also a precursor to MBOS. It already includes elements of the MBOS architecture, and it brings our customers some fantastic new features. Ladies and gentlemen, it's my great pleasure to present to you the interior of our brand-new E-Class. [Presentation]

So isn't it the beauty, intelligence and digital worlds are coming together. So earlier, we gave you a deep dive into our high-performance infotainment computer, and we have pulled it forward into the E-Class. A variant of it already runs interior systems such as infotainment, instrument cluster and cameras. So this differs from approach -- previous approach of multiple control units in a car. It enables us to create an immersive entertainment experience that is more interactive than ever. Our new MBUX API for Android compatibility layer in the E-Class makes it easy, right, to install third-party apps. It offers a far better user experience than with mirrored apps. For example, you can run a Webex or a Zoom call while the car is parked. The centerpiece of the digital experience inside the new E-Class is our new MBUX Superscreen. A large glass surface extends across the entire width of the vehicle. It seamlessly integrates a central touch display and passenger touchscreen. The new E-Class is a stunning combination of luxurious home and digital office. And many new features allow customers to create their very own personal space. For example, our routines feature is completely new in the car. It enables customers to configure their E-Class for their own everyday needs. You can create and add a wide variety of routines on the main display. Lots of combinations are possible -- for instance, switch on the heat seating and heating and the ambient lighting to warm orange every time the cabin temperature is less than 12 degrees. Thanks to the over-day updates, the new E-Class and its MBUX interface, we get even smarter with time, unlocking more and more benefits for our customer. And on that note, Britta,the floor is yours. Please.

Markus, thank you very much. Warm welcome from my side. So all of these digital developments that you've just shown have just been designed to fulfill a mutual goal we share with our customer, which is creating the best experience in our living space. So our goal is to build a house where our customers can truly feel at home in with software-enabled upgrades that are continuously evolving over time, our customers can fully adapt their Mercedes-Benz living space to their needs. In everything we do, our focus is on making the Mercedes-Benz digital house easily accessible where our customers love to say. So let's have a look. [Presentation]

So when we ask customers about the importance of digital services, more than 80% said, digital extras are very important. And that's reflected across all markets and all demographics. So the digital extras play a fundamental role in the decision to purchase a car. And our customers have high expectations about their Mercedes-Benz digital experience. They want it to be simple, intuitive and functional. And how do we know this? I mean by utilizing data-driven feedback in a real time with over 5 billion data points per year. And this is how we gain a better understanding of how our customers use our products. Moreover, they expect their vehicles to be always up to date, thanks to regular over-the-air updates. And the Mercedes me ID is both, the basis of our house and the key to accessing the Mercedes-Benz living space. It is a unique customer ID that gives us a strong and direct connection with our customers. Easily accessible means just a few clicks. After registering for Mercedes me ID you need to accept our terms of use, enter your pin and activate the services. So our customers have the choice of which services they want to use and which data they want to share. Personal settings and profiles gives customers a high level of individualization. And as mentioned before, privacy by design means privacy and data protection are prime focuses. We already have a strong customer base, which we are building upon to further expand and improve our offers. So Mercedes me is currently live in 50 markets. In total, that represents a network of more than 10 million connected cars worldwide. As mentioned already, by 2025, we aim to expand to over 65 markets, and our fleet of connected cars will grow to more than 16 million. That would represent 60% growth in only 2 years. With over 15 million downloads since its launch, our customers love using the Mercedes me app, making it the best rated app in the industry with 4.8 stars. On top of great customer feedback, J.D. Power also awarded us as having the best app among automotive OEMs. The app offers far more functionalities than just remote control. It's also a tool for communication and other customer engagement and sales. So looking at sales, we had 60% of 1.5 million Mercedes me store purchases last year already made via the app, which means an increase of 87% compared to the previous year. So our customers enjoy the high flexibility in choosing how, when and where they want to access the Mercedes me store, whether via the smartphone app or our web store or even directly in car, our customers have easy access to our stores. In-car purchases are just as easy using the Mercedes Pay fingerprint feature. And we are simplifying the customer churn even further with our Mercedes virtual assistant, leveraging the latest conversational AI capabilities within the Azure OpenAI service, we are able to deliver best-in-class customer service along the entire digital customer journey, starting the end of March 2023. However, any store can only be successful if it provides great offers. And as you just saw in Markus and Magnus' presentations, we have already a lot of great features today. For example, within our entertainment portfolio, our customers can enjoy the most popular streaming services such as Apple Music, Spotify or Amazon Music. And we are well aware that local services are highly popular, which is why we offer entertainment features that are tailored to local and regional markets. So for example, Tencent Video, QQ Music and Ximalaya are market-specific offers in China. Additionally, we're the first carmaker that has integrated ZYNC in-car video streaming, giving our customers a growing selection of movies, TV series and more to choose from as we just heard. And of course, ZYNC also offers the regional content. ZYNC will also be available as an over-the-air update for selected models later in 2023. And that's just the start of where we are going with entertainment. On top of that, Guard 360 includes services that enable the remote protection and monitoring of your car. These were just a few examples of our wide range of attractive and exciting software-enabled upgrades. We have continuously expanded our offers to date, including remote, navigation and entertainment products and services. These product selection will continue to grow, but what we offer won't grow in complexity. We are giving customers a simple, intuitive and transparent experience. So starting with the launch of MBOS. We will unite all our digital extras into 1 single packaged called MB.CONNECT. -- for maximum ease, a generous data plan and regular over-the-air updates are also included. We'll offer flexible contract options so customers can choose to activate their MB.CONNECT subscription at the time of purchase or at any later date. In the MB.CONNECT environment, we anticipate greater than 80% customer retention for vehicles in the 1 to 6 year age. In addition to MB.CONNECT, we will expand our offer for charging and automated driving with MB.CHARGE and MB.DRIVE. Let's first have a look of what we offer with MB.CHARGE. With 3 MB.CHARGE packages, our European customers can currently choose from flexible, transparent and attractive offers depending on how much they drive. And according to our forecast, we expect more than 80% of all customers to use 1 of these 3 MB.CHARGE packages in the future. MB.DRIVE will incorporate assisted driving, enhanced assisted driving and conditionally automated driving. It's expected to become our largest revenue driver in the software-enabled environment. We have set ourselves ambitious targets. So starting with our MMA platform, we aim to make all new models hardware ready to enable assisted driving or enhanced assisted driving functionalities. Selected models can even be ordered with conditionally automated driving with a fixed-term contract from the factory. All driving levels will be continuously improved via over-the-air updates as it has been illustrated. As we've mentioned earlier, our connected car park is forecasted to grow by 60% in the next 3 years. This will have a huge impact on our software-enabled revenue. In 2022, digital extras generated revenue of more than EUR 1 billion. MB.CONNECT fixtures comprise the largest software-enabled revenue share in last year. By 2025, we are targeting more than EUR 2 billion from our connected car park. In addition, automated driving functions are expected to emerge as our largest software-enabled revenue driver by 2030. And the digital transformation in sales and marketing goes even beyond. So as outlined, software-enabled upgrades are an important pillar for us and for our customers. The online to offline experience is essential to a growing e-commerce environment. And the fast rollout of our direct sales model will strengthen our overall customer interaction. As of now, we already have 7 markets with the U.K. having gone live January 2023 and 3 more markets to come later this year. As a consequence, we are working on optimizing our retail footprint to be ready and effective for the future. All of this will make quite an important impact on our future business. And with this, I will hand it over to Harald who will give us an outlook on the key financial aspects. Harald thank you.

Yes. Hello, and welcome from my side. You have seen 3 packages for our software-enabled main upgrades in Britta's part, The MB.CONNECT, the MB.CHARGE and the MB.DRIVE. I will now dive a bit deeper into the revenues from MB.CONNECT and MB.DRIVE. I would not talk about MB.CHARGE today. Since the majority of MB.CHARGE is part of a separate unit within the recently announced high-power charging network. We will talk about that at a later point in time. First, let me talk about MB.CONNECT. Today, we have already great features available generating software-enabled revenues such as navigation services, online map updates, live traffic, vehicle monitoring system. Most of today's features are embedded in our vehicle sales and margins. The customer sees the benefit and the pricing is embedded in the overall vehicle with finite durations. These -- with these embedded main revenues, we think we have achieved approximately a bit over EUR 1 billion revenues in 2022. These revenues grew slightly, but are very stable over the years with decent margins embedded in vehicles margins and revenue. Now there are great things coming on top. For example, music and video streaming like ZYNC, features like Webex and Zoom or on demand like dashcam or MBUX augmented reality. On top revenues, we will offer both ways as initial vehicle factory sales and as store sales. In case, I mean the customer chooses factory sales features come with different durations ranging from months up to several years. At the same time, any point thereafter, we offer it as a onetime purchase or subscription with renewal options as the so-called store sales, e.g. as monthly subscription. These features have a price tag and can generate recurring revenues over life cycle. We think these features are very attractive and offer substantial benefit to the customer. So the customer does not want to miss that. That's why we expect a high take rate when the vehicle is sold from the factory with a high renewal rate in the following years. Main part of these features starts with MMA and keeps growing with MBA. We expect with the on top to generate less than EUR 1 billion revenues by the mid of the decade and more than EUR 1 billion of annual revenues in the second half of the decade. We also expect healthy margins on these. It remains to be seen, however, which way the majority of customers prefer the purchase via factory or the purchase via store. Both is fine for us. I said it before, if customers may choose factory sales, it is part of vehicle profit. Only if they choose to buy them later, it will be incremental profit such as profits on subscription schemes. But the customer behavior is probably a bit difficult to predict at this stage. So MB.CONNECT, we think that total revenues out of Embedded and on top will add up to more than EUR 2 billion in the second half of the decade. Now let's have a look at the MB.DRIVE and our partner for that. Before we look at the revenues of MB.DRIVE, let me walk you through the business model with NVIDIA. We have entered the partnership with NVIDIA because we're convinced this is an intelligent and an economically efficient way to deliver world-class ADAS capability. We joined forces with an extraordinary capability technical partner, while at Mercedes, we are already leaders in ADAS as our L3 drive pilot attests. Within NVIDIA, we have access to software development, AI and processing skills that will take us to the next level as Markus, Magnus and Jensen explained already. We are happy to have them as a trusted partner in a win-win relationship. NVIDIA puts focus on AD, based software, software updates and SoC integration as well as the back end of the system. We put focus on vehicle integration, application development and vehicle equipment. 2 powerhouses with different roads, joining forces and thereby speeding up development times. By that, both of us enjoy financial benefits of the corporation in a fair and balanced way giving us a profitable ADAS business. How does it work? Each partner bears its own cost and investments. NVIDIA has a majority of upfront investments, especially on the R&D side. We equip the vehicles with our Level 2 and Level 2 only features and carry the respective hardware costs, including favorable SoC pricings from NVIDIA. Due to the high attractiveness, we assume a high take rate for these features. In the end, we, as Mercedes, generate 100% net sales and share roughly half of it with NVIDIA depending on the region. Even by doing so, we still expect a decent contribution margin from this business, but lower than the one on our MB.CONNECT. However, L3 features are only available on demand, meaning only as an option. Therefore, we have no risk in terms of take rates. So you see this is an intelligent and equitable partnership. So what does that mean for our MB.DRIVE revenues? The structure of these revenues depends on what the customer chooses. So it depends whether they go for factory or store purchases. Clearly, we expect it to ramp up with MMA and later MBA to a low single-digit billion euro figure by the mid of the decade. However, with the scaling of the portfolio, we expect this to grow to a mid-single-digit billion revenue figure later in the decade. Depending on the customer choice, factory or store sales, the results will be visible as part of either our vehicle margin or separately as subscription proceeds. Now let's have a look at both together, MB.CONNECT and MB.DRIVE revenues. First, we have the embedded revenues with around EUR 1 billion in 2022. Second, we have the MB.CONNECT revenues coming on top. And third, the MB.DRIVE revenues. All in all, we expect roughly a low to mid single-digit billion euro figure by the mid of the decade. This, we expect to evolve to a high single-digit billion figure by the end of the decade. On EBIT, when looking at the mid of the decade and what we have in our embedded, CONNECT and DRIVE, we expect definitely more than EUR 1 billion EBIT in 2025 altogether. However, as I also mentioned before, at current point in time, we can say if we will see it as part of our vehicle sales or separately, it depends on customer preferences. With all of these things, I said before, we expect these amounts to grow further in the years ahead. We are on track versus our 2025 target, and all of the above is already part of our profit tables and weather chart ambitions. What are we doing in the invest side? We rebuilt our house within the financial framing we gave ourselves. Therefore, the investment needed for the transformation are given priority. Looking at our R&D investments now, we see a significant step-up and an increased share of EV and software-related spending. The R&D cash spending for our MBOS software and core spending hardware is at a run rate of roughly EUR 1 billion to EUR 2 billion a year. This investment is for our Mercedes-Benz operating system, where the largest amount is accounted for in the infotainment and in the ADAS integration. We plan to increase the software share to around 25% of our R&D investment by the mid of the decade. At the same time, we reduced the -- in the investment in ICE as any new platform development will be BEV only. Total invest of R&D and industrial CapEx will be reduced by 20% versus 2019 by the mid of the decade as planned, but don't nail me down whether it's 2025 or 2026. Let me summarize. We don't tell you fantastic stories how software-enabled revenues will reshape the business model completely. We just told you something more prudent, but not less important. We are focused on one thing: offering the world's most sophisticated way to move from A to B. combining luxury, and tech is the main part of the Mercedes sophistication and a key element to spark desire. It supports our luxury position of the brand and the respective pricing of the vehicles. We create outstanding features, making our products mean even more attractive. Thereby, MBOS is an important enabler for our strategy as well as for our margin ambition. MBOS supports pricing and margin. And yes, it has a potential for additional revenues and margins. At the same time, we're doing it within existing envelopes in a fair risk reward sharing model. But maybe Ola you see even a bit more potential than me.

Yes. Thank you, Harald. We have now come to the conclusion of this presentation. I just want to sum up some of the insights from our point of view, what is this all about? I started with talking about -- it is about our strategy, our ambition to build the world's most desirable cars. But what we have seen today with MBOS is really an opportunity. It's an opportunity to create outstanding products, create stickiness with Mercedes, create revenue opportunities, and it goes beyond the product and the product sphere. It also gives us a chance to improve the cost situation, productivity, speed, time to market on the enterprise side. Our conviction is software is a core competence for a car company. It's a core competence for a car company. And we are building this core competence and have been building it over several years now. But we're also realists. You cannot do everything yourself, nor should you do everything yourself. It would not be the quickest and not the most cost-efficient way of implementing a complicated system like MBOS. So we partner and we partner with the best. If we get that right, what are we really doing? We are future-proofing the valuable real estate of our vehicle. The interface, the one-to-one connection with our customer. And that is then underpinning the revenue opportunity that Harald was talking about. So that's what it's all about. That is MBOS in a nutshell. But I would like to mention the last piece and no pressure here, Markus and Magnus, but now we have to deliver. I know we have the E-Class that is coming this year, and it was no small feat to get that up and running, but we have to deliver this whole thing. So we are now laser focused in the next 18, 24 months to get this journey to its 1.0 stage with the MMA architecture that we will present towards the end of 2024 and go into the market in 2025. Thank you for listening. Thank you for coming today. And this is just the beginning of a journey. Thank you very much. [Break]

Thank you, everyone, for your presentations before. And now welcome back to our analyst and investor Q&A. Ladies and gentlemen, a warm welcome to everyone who could join us here in person in Sunnyvale, California and also to everyone who is following us on the internet. That's quite a few. My name is Steffen Hoffmann, heading Treasury and Investor Relations for Mercedes. We are very happy to have all the speakers with us, Ola, Markus, Magnus, Britta, Harald. Great that we are here for you for the Q&A. I'd like to remind you that presentations and Q&A are governed by the so-called safe harbor wording that you'll find in our published documents. So what is this all about? Please note that the presentations and comments contain forward-looking statements that reflect management's current views with respect to future events. Such statements and you know that are subject to many risks and uncertainties. If the assumptions underlying any of these statements prove incorrect, then obviously, actual results may be materially different from those expressed or implied by such statements. Forward-looking statements speak only to the date on which they are made. For the Q&A, we will answer questions from the audience here in this room. Those who are following us on the Internet are invited to send their questions via the MeetYou chat function. You can find the link in the IR release that was sent to you yesterday. [Operator Instructions] We start with Tim Rokossa from Deutsche Bank.

Great presentation, I have to say, really like the condensed format, but also how you explained things that were very technical in particular from you all. So for none of us who really understand so much of that. So I'd like to start with 2 questions, please. The first one is a little bit two-folded. On the software revenues, just to understand your assumptions, right? I now pay more than EUR 70,000 for Mercedes. You very successfully increased your price, Harald, you know, we like that. But it also means as a customer, I would expect members is already to be able to do quite a bit for that price point. So when you do your calculations, how do you determine, Britta, maybe to you, or Ola to you, what are customers still willing to pay for on top of that price that they have anyway done, and don't they just expect the vehicle to always have the latest and greatest and remain very fresh. And related to that also, Markus, perhaps to you, all of your partners here today, great to have them all here, a big respect for you, certainly, but they all made it very clear that you were only their first partner and the lead partner maybe, but they will roll this out with other guys as well. How much of a technological edge gives you this over 1 year, 2 years, 3 years, what should we expect? And then as a second question block, can we -- we found ourselves a little bit in this subset situation this morning that we were trying to find a traffic jam to actually learn how the Level 3 feeling is when you talk about 130 kilometers driving and all these type of things, is that something that's going to be very near term? Is this 2025, 2026? Or are there still so many corner cases that you have to fix that this will remain an ambition for a few more years to come? Or is it really applicable immediately and not just if we're trying to hand for traffic jam.

Britta, do you want to start?

Yes. Basically, what we see, and this is why we flexibilize our offer. The customer want us that we flexibilize for him the possibility to choose. As of now, we are -- we have this embedded. But what we are seeing as of now is customers as well asking for more flexible solution, how to use the car. Not everybody is buying it for a lifetime. Not everybody is asking in the future to lease it for 3 years. We are increasingly offering more subscription model for cars as well. And with this wish for flexibilitation, this is as well going on the -- because you are referring to the MB.CONNECT package, obviously. So we are flexibilizing our offer, which still would bring us into the position that by going to market, we are still free to decide what is then the right go-to-market for which model, and the nice thing is that we are flexible, we can give the flexibility to the customers but as well to us.

Markus?

Just to the question of the partners. We are in a company that lives off innovation. And that's a big pillar of this company to innovate first, to be first in the market and to drive innovation with partners. I mean this has been our DNA and it's going to be our DNA. Now in the digital space, particularly, and that's why we're working with these partners here, developing new products, not taking existing product. So we're working on the next level. So enhancing what map and navigation is in the future with these partners. But our interest also is that they are able to scale. So that's absolutely okay that we start with them and develop a product, be first in the market and have disadvantage for quite some time, but also our interest is the scale, so that they're going to scale over time. And we're going to move to the next level here.

And of course, building on standards, new innovations can easily then be added on at speed, and that's the importance, right? So being first and then continue to innovate adding new features at speed.

But that concept team is not new. You can go back decades in history, I don't know, from inventing the airbag or ABS or whatever, we do something. It then some years later become perhaps even an industry standard. I mean it's the same thing here. But when it's an industry standard, we're on to the next thing.

Well, we built partnerships, but what we explained today, we built the core system, and then we pick and choose. So we have our interfaces where we pick and choose the best and customer preference is changing also as we see, and it's different from region to region. So this gives us flexibility also over time to stay flexible there. Then your question on 130 kilometers. So I would go back to the current system, the DRIVE PILOT. And one of the news today was the increasing speed and we said we're always gradually going to improve the system, safely. And safety is the utmost importance here. So we're going then from 60 kilometers that we have today, we feel confident that we can go to higher speeds with the current system today. So we have some confidence in what we are doing here, taking the speed up from 60 to around 90 kilometers around next year. So this gives you an idea that we -- judgment and have a judgment what happens if you go to higher speeds up to 130. So we see this as possible. It needs compute power. It needs latency that we go down, the bandwidth and all what Magnus explained here, and it need great sensor technology, and we need new sensor technology actually to look just further down the road, 600 meters capability of a LiDAR that's not existing today. So today's LiDAR can't do that, but future technologies of sensoring can do that. So we feel confident as we're going to start with MMA and M Beam and then we introduce in the next step than the higher speeds of Level 3.

We haven't nailed down a specific date, but it's not like we're putting it so way out that it's not relevant. It's a serious development projects.

We continue with Horst Schneider from Bank of America.

Yes. I want to come back to the Slide 7 from Harald, where you have shown all the revenue potentials that you want to achieve. You said also that the margins on MB.CONNECT is higher than on MB.DRIVE. My understanding is the reason why other car makers didn't want to take NVIDIA as a partner as they had to share revenues. So therefore, can you maybe shed some more color on what is the margin potentially you want to achieve from that? And putting now this new revenue potential into the context of the 2025 targets, 2025 is more a midterm target. I ask myself, what is the long-term target potentially? I don't expect you now to give a precise margin target, but just to say, is the margin then more flattish? Or is it up? We talked about that last night that you said BEV profitability is still not as great as it could be maybe, so what's the path from 2025 to 2030. And then again, coming back on this Level 4 driving since you don't want to commit to the specific date. I remember from Porche, they said, they do not want to offer that before 2030 or they're not able to offer that before 2030. Would you agree to that for Mercedes as well?

So on the revenue and the margin. I think -- when I explained, I mean, the business model with NVIDIA it became clear that I mean, there's a fair sharing as I mean, they have really been the lion's share of the investment upfront. I think it's only logical and fair to share also, I mean, the benefits from it over the time. And that means, obviously, that mean once you do that, I mean, the margin, I mean, it's the lower one compared, I mean, to the CONNECT, where basically we have more of the vertical side of things and therefore, have more margin. So I think that is self-explanatory. I mean, as such. I mean what is the margin exactly maybe I keep lips a little bit tight on that one. Anyhow, I mean, what I think we explained here today is that you need to see altogether, I mean the embedded revenues on the CONNECT on top and the Drive. And all of that, I mean, together, I mean, we confirm. I think you could see it as well more than EUR 1 billion by 2025. But frankly, we don't know today, as Britta said before, we offer the optionality to the customer. Is a part of the vehicle price at factory. Is a part of ex-tor later. I mean, is it approaches ex-tor? Is it a subscription? So we give the full flex. You guys want to nail us down then later, I mean, for sure and see where is this line in the P&L? No. I mean, it will be either. I mean, in the vehicle margin or it will be, I mean, in a separate item. And that's why we give it as a holistic perspective, I mean here today. Well, when we say it's more than EUR 1 billion by the middle of decade, I think I said it should grow, obviously, by 2025. If we are on the revenue side by mid-single digit billion number and towards the end of the decade, I mean, a high single digit. Probably there is a case that the EBIT will grow a bit beyond the EUR 1 billion. But the key point here, I think, today is much more it supports, I mean, the strategy, it supports the positioning of the product and how we're going to monetize, we leave that to the customers. But I think what we can take away here is there's a great potential from doing so.

But let me help you Horst with your spreadsheet because we know how to do spreadsheets as well. So I'm going to be 1 more notch specific. Do you do a risk and reward partnership or do you do a classic just buy technology and pay the full price partnership. Let's start with that, which is a well-known model. In the well-known model, you pay a significant amount of supplier development cost. You have your own cost. The supplier has all of his other base development cost and the other fixed costs associated with the business model. And of course, in this case, it's hardware and software. So you pay the full price in that case if you would go to a classic of the chip. That number is a very sizable number. We pay none of that, nothing and only pay a heavily subsidized chip. But then together, figure out how do we maximize our joint revenue. The reasons why the margin is lower than on the CONNECT side is that if you still add up everything, this is the most sophisticated technology in the car by a country mile is that the total of the variable cost that goes into that and the amortized fixed cost is higher than showing a movie in the car. And unless you think you can have unreasonable pricing, but I think the whole point of Harald's presentation was not to go to pie in the sky type of pricing assumptions, but I still think that there is some kind of an equilibrium here then automatically, mathematically, you have a little bit lesser margins than you would have on a CONNECT feature. So we have thought through this very carefully.

On the question, so I understood why we are not more ambitious. I think we are pretty, pretty ambitious when it comes to automated driving. How many companies that are out there today that have Level 3 it's a huge problem to solve because that's why there is only one company that has a Level 3 certification at least in Germany and for Europe, for 48 countries, that's Mercedes-Benz currently. So it's a huge problem to solve, and we want to solve it safely. And safety is always in the forefront of what we are doing here. And I think it's pretty ambitious to commit to Level 3 with our next generation of cars based on the MBOS system here and going to 130 on Level 3, that's pretty ambitious. The Level 4 was the parking scenario, which we have obtained also as the first company. And that's, of course, we're going to transfer also in the future. But all in all, to do it safely, I think a huge ambition.

So to elaborate on the Level 3. So today, we are the only company that has a Level 3 certification and this is on the highway, the autobahn. So what we're doing right now is to using machine learning and AI in order to solve that use case in the urban environment. And it's a very, very difficult problem to solve on Level 3 and then using this partnership. Others are doing fantastic sake, but that's Level 2 systems. It's a very different thing.

Just a small add-on, maybe, yes. I didn't want to say that you are not ambitious. You are very ambitious, but just the revenue potential you've outlined does not include at all Level 4 driving, correct?

No, if your definition of Level 4 is true Level 4, I'm going to make an example for you. You are asleep in your car, while your car drives you from Stuttgart to Hamburg. That would be Level 4. That is not included.

Okay. So that's automated, that means it's beyond 2030, right?

I didn't say. I just said it was...

So we work ourselves to roll through. So we continue with Tom Narayan from RBC.

Yes. So Ola in Monaco, you said you kind of set up the event today to prove how owning your own operating system, why that's important. And I think you guys did a great job, obviously, today. You're 1 of a few -- a handful of automakers really seeming to doing it in-house, let's say, with partners, but really in-house. Some others are using third parties to completely outsource ADAS capabilities. The question is, and I don't know if you agree with this premise, longer term as some of these features become mandatory, right, just safety, like you mentioned, airbags or rearview cameras, what if a lot of these features just become mandatory, wouldn't the pricing come down that you could charge your consumers. And in that sense, would it not make sense to outsource completely ADAS capabilities to somebody else like some other automakers are doing. And then second question is on Google Cloud. It's kind of the opposite of the first question. By doing that, presumably, yes, other automakers would also use that. Does that, in any way, harm the Mercedes brand or not harm the Mercedes brand, but you guys want to differentiate the experience. Does that in any way impair the ability to differentiate the driving experience.

So on the first point, I think, especially in the presentation of Markus & Magnus, we made it absolutely clear that MBOS is so much more than just one of the domains. And quite often, people when they talk about an operating system, they just jump on to one domain and talk about that specific domain, but it's not. And the full integration effort both on the hardware side and on the software side. It's not something you can buy from anybody. The OEM is not going to be able to release that responsibility of the full integration. So in one way or another, you're going to be involved. We have just chosen to make sure that we -- with very clear, let's say, technological strategy and rules have figured out how we're going to manage this. Inside the domains, you can then decide partnerships. You asked specifically about automated driving. Yes, there's really the full spectrum here. Some car companies are trying to do everything themselves, like literally every single bit of it. Some car companies, they go out and see what can you buy? Let's see, if I buy that, but they still have to do the vehicle integration and the application of it. We have chosen, in our point of view, kind of the best of both worlds. We have 20-plus years of experience in assisted driving. We know a lot. We know a lot about both the hardware, the software, the application and everything. But you really need a compute partner that is the best you can find in our view, NVIDIA and with our software capabilities in the business model that we have just described. So we think that what we have chosen for us is actually the fastest an economically quite intelligent way of solving this problem. And we'll take it from there.

Yes. So in MBOS, I think the key is to combine the different domains to a whole new experience. And it's not only what one domain can do, what ADAS can do. But to combine several domains to a new experience, that's the trick actually here. So the combining ADAS with the DRIVE domain, for example, and creating a new experience for the customers or additional benefit. So I know what's going on in my drive train while I'll navigate or I do assistance using the assistance or combining the infotainment with a navigation to combine -- have a new combined experience for the customer, which will show actually in the near future. So just being able to do that, to put this all together and coming up with a new product, and we are in the driver seat, and that's the beauty of MBOS.

Regarding the second question on Google Cloud, we want Google to offer this new product that they're coming out with to the mass market because it gives just a larger user space that gives us better data. for the Places situation. And then what's so excited about is that Google coming out with this premium product accounts now we can really start to innovate. We can start to see how does that put together with our 3D game in [indiscernible], how does that work together with the other domains, how can we elaborate this. And this is something that only we can do.

And we just go to the next gentleman, George Galliers from Goldman Sachs.

Thank you for the very insightful presentations earlier. I had 2 questions. The first one was really just on the standardization little bit you mentioned with respect to the sensor set compute on the vehicles, starting with MMA. Does that include everything from the LiDAR through to Ultrasonics, or will there still be some differentiation between vehicles that are equipped for Level 2 versus those which are capable of Level 3. The second question I had was really just on MB.DRIVE and MB.CONNECT. Just to be clear, if as a customer you subscribe to these, did those subscriptions belong to you as a customer and you can, therefore, use that functionality in any Mercedes with the appropriate hardware and software? Or do they belong to the vehicle, and which do you actually think is the bigger addressable market attaching it to individuals or attaching it to the cars?

Should I start with the second one briefly? Very clear, the license days and is with the car because it's the feature that we need to authorize via the mic for the car. So it stays with the car because it's as well depending on what features do you have in the car and what features can you buy for the car. So it stays with the car and the customer buys it and the license for the car.

So with regard to the sensor set, the -- I'm going to call it base sensor set for Level 2+, as I mentioned in my part of the presentation, in terms of how met sensing technology we're putting in the car. It sits above the most sophisticated assistance system that you can buy in the S class today, bar, the Level 3, sorry, bar the Level 3. And that will be for every single Mercedes then going forward from that point, including the supercomputer, but Level 3 where you have the LiDAR and even more of the other sensors, that is a high version that is an option that we're not going to put in every single car because the variable cost of that is very high.

Henning Cosman from Barclays.

We -- or I at least come away thinking you're really at the forefront of a lot of the things that you talk about, right? And therefore, I'm almost asking myself, are you almost depriving yourself of an opportunity with a use case where you have obviously a luxury, premium luxury manufacturer, whether with limited annual sales and a growing but limited fleet size. And is there almost a scenario where you could think about licensing or your significant development effort and being at the forefront of that technology, perhaps making that available to car manufacturers with significantly higher volumes than yourself that you're perhaps not directly competing with. Is that in any way an opportunity, or why would it not be an opportunity considering that you yourself are also investing a lot in these things? Or would that opportunity fall to your partner, somebody asked before you're going to be the first partner. There's going to be other partners. When are you losing your edge is there perhaps an opportunity for you to leverage your own investment costs into licensing opportunities going forward. That's the first question. And then I don't know if I'm going to get an answer, but just to take advantage of having Britta on the stage as well. Horst asked about the margin trajectory in the context of the additional revenue opportunities, margin opportunities offsetting some of the battery electric dilution. We didn't talk much about the direct sales model today, but that is, of course, something that, for sure, qualitatively, you're thinking of having an offsetting, positively offsetting margin effect as well. So because we have you and we don't have you that often. Could you help us quantify that opportunity perhaps as much as you can from today's perspective.

On your first question, it's not something that is on the cards today. And I have to admit for these guys that have to do the work and the thousands of people behind them that do the work. To enter yet another cook into the kitchen would just complicate things. It is so sophisticated that for these next few years, we will definitely concentrate on ourselves. Whereas one should never say never, philosophically though, if I think about your question, it's nothing less than the brains in the central nervous system of your vehicle. And that is the core of your vehicle. I find it in a way difficult just to comprehend, at least from our philosophy that you would buy the brain and the central nervous system from another OEM. I'm not going to say that it's not going to happen, but at least for us, Mercedes people, that feels strange.

But we might give it to the van people, right?

Yes, yes, the van, that's us. And we're not saying anything about the pricing here, right?

Yes. We basically let's -- just a second, and Steffen and myself, we had yesterday a discussion. So maybe I would be prepared to go even in deeper discussions about all the transformation and sales and marketing if you want to, but just trying to give a short answer. So the overall starting point for doing the direct sales was controlling the pricing, when you go more and more digital because if every dealer would go digital, then you have the full transparency. So if you have the strong belief that digital sales will become more of an -- of essence for our customers going into the future. This was the starting point. Then we said, okay, obviously, there are 2 other big levers you control because that's the key. As a result, you control the price because your formerly dealer is now an agent, so you can set the price. What we see currently as we are running both models in parallel. In the recent 2 years where we have seen our current dealer model versus the direct sales model is that obviously, when you control your own price, we have -- we had a better perspective on how we control the price than if you are doing this via the dealer. So this was the first very positive effect. Some of these effects are built in the weather chart already with the forecast. I still do believe there is much more that we can see over time when we are getting now into the bigger markets when we are adding more markets in Europe because the efficiency in the model, we are just starting to see, so I think we are still at the beginning, and I think it's the right way to go, and it's one of the key levers of sales and marketing to -- which is built in the weather chart of Harald.

Yes. But let me come back to the weather chart shortly. And you know that basically the original of the weather chart is a transition into BEV. And I think we all know the challenge of going there with structurally lower margins given the high variable cost on the BEV vehicle. And we put together, I think, a comprehensive plan mean another part of which you saw here today, how we're going to address that challenge. So what I mean, we saw today, what you -- when you heard in terms of the margins, I mean, coming from CONNECT or coming from DRIVE, I mean, the benefits coming from direct sales, I mean the benefit coming from digital sales. All of that is part to manage the challenge of the transformation going to BEV and therefore, is completely embedded, I mean, into the weather chart. I mean, this is really an extremely important, I think, for us that we are all on the same page here.

Harald Hendrikse from Morgan Stanley.

That's the most comprehensive presentation we've had on these technologies from any legacy OEM. Two quick questions. Firstly, you talk about putting the compute and chipset into every car, again, sounds like the right decision, but it's also a brave decision, right, because you don't really know the revenue profile of the investment you're going to have to make. How are you going to deal with the extra cost of putting those chipsets and computing power and the sensors in all of the cars prior to getting any of those revenues. Is that going to be just the cost of manufacturing? Or is it going to be sort of seem to be an investment CapEx type of spend? And then the second question, sorry to go back to Google, but not specific to you, but the German industry, I don't think has been super keen to get Google into the car from day 1 on this whole project. I think your decision is practical, and I think it supports the strength of the car and the brand in the future. So I like the decision. But can you maybe talk us through the thought process where we've gone from -- I feel like we've gone from a situation where Google is never going to go into a German car and now it's going to go into your car. Maybe talk a little bit through how you made that decision, please?

On the first point, here comes the beauty of the business model. If you are not in an all buy scenario, but you actually are in a -- both parties are interested in maximizing the revenue and the profit of the thing. And as we hinted to, even though we're not going to give you the exact numbers, if you get that supercomputer at a heavily discounted price, actually, the variable penalty issued the customer not turn on any feature is not that high. So it's not an investment. Yes, it sits there in the variable cost, but you make an economic model around how high do you think the take rate will be of the different offerings that you have, plus the opportunity to enhance it over the life time of the vehicle and upsell in combination with which price you think you can charge for the different levels of service. So it's a changing business model. And I realize it's a bet. It's an unusual bet compared to how we have been doing business for a long, long time. I mean it's no secret that prominent start-ups started with that bet from the word go. And the more we thought about it. We said this is a bet we want to make because it sends another message. Every single Mercedes has a capability in terms of its assist and safety that goes way beyond where we are today. So even if you are the customer in the beginning and you're maybe very financial oriented, you don't buy it, the guy that buys the car off of you, as Britta said, could you just turn on that feature later on. So we think this lifts the brand. We have always been about safety. And you get a super comprehensive sensor set with a supercomputer. Isn't that fantastic? Starting with the entry vehicle, I mean, it's crazy good for the customer.

Well, to the second one, maybe I'll start and then I hand over to Magnus. It's all about customer experience, right? That's what we said. And we're building the core of the system and then we plug and play the best players, the local heroes, wherever they are in the world. And that's what we're doing in China with AMAP, and we went to AMAP there because this is the most trusted map provider in China. If you look at points of interest today in our current system, yes, you get some nice point of interest, but it's not what a smartphone customer experiences today when they are looking for point of interest and look for great ratings, a lot of people judging about a place a restaurant or something. And that's what Google provides definitely. That's the most trusted source in the Western world. That's what it is. We are not implementing GAS, Google Automotive Service in our cars, so there's a change to the business model and that's the result of our discussions with Google that we said we need to come up with a different model, not the traditional GAS model that they're offering today. So we were sitting together, putting our heads together and developing a new product that's different from GAS and that we plug-and-play into our system, but maybe you talk about how we do it.

Yes, that's right. So having a dialogue with the tech community and being here in the Valley, having those discussions currently, when we're then talking about the vision, what we are trying to create, we're trying to create this digital luxury and what does it mean? And how could you do that? I mean having that dialogue, it became clear in the dialogue we were having that, okay, this needs to be a new type of offering that can elevate and broaden the portfolio of Google, they talk about that in the talks we're having jointly later. But for us, that opened up completely new worlds because all of a sudden, we have a partner that we can innovate with, we can continue to improve. So when we talk to partners, when we talk to all kinds of people out there, we're not looking for vanilla. We're not looking for so, we want to look at tech partners that we can together create something new and unique for our customers. That's what digital luxury means for us.

So we have a question here. Very close to the camera.

Can you hear me?

Yes.

My question was about the Drive, the Mercedes-Benz Drive. You included in the revenue share base software. I was just curious where the line is for that. Who brings parts of the Drive software to the table in the perception part, localization part, those things. Where do you draw the line where it's Mercedes versus NVIDIA?

Well, maybe I jump in. Maybe on the financial side of things?

You can't do that in financial terms. It's one software that we have codeveloped most of so, yes.

So all the components are codeveloped?

Not all the components, but many of them.

Got it. And just curious as to coming to market, some people are coming to market with software like Mobileye or the Qualcomm are going to come to market with Driver software that was kind of independently developed. What are the components that are not yet developed here within the Driver stack that still have to be kind of either acquired or developed between yourselves and NVIDIA?

Okay, I can answer that one. So Sarah and I have a very tight joint team that we call Nexteam that we're working together to bring it in. And the issue is -- the problem statement is really how can we do this safely. So how can we prove this so that it is basically living up to the Mercedes pledge of always having a safe approach. And to prove this, you basically need a lot of data, as we said, and we need to be able to run through all of our models and the simulations and to do this. So basically, it's hard on the system to make sure that the system is there. You saw the video. It's driving. It's just doing what it's supposed to do, but making sure that this is safe, this is the challenge ahead.

Last question under that. I appreciate that. Is the kind of the goal, one of my colleagues here asked about Level 4 and you said, look, that's -- we're not talking about Level 4 here. But with regards to Level 2, is the target to deliver by the end of the decade? I think that's where the timing was. Level 3, eyes off urban type of driving? Is that -- would that be the appropriate target for that?

Level 3, urban, that's the target. Absolutely.

Thank you very much. We continue with a question that reached us via the MeetYou webcast app. From Jose Asumendi from JPMorgan. I think it goes Markus to you. First of all, he's thanking the team for the great presentations. The question is about ADAS, Level 2, Level 3. How far did you get in terms of millions of miles driven, number of disengagements on the highway? And how do you think Mercedes compares against peers, so he refers to Tesla in ADAS experience as the car becomes autonomous and the driver wins that time in the car, which functions do you think drivers will perform in the car? And how large is the business opportunity.

Going, again, back to what we said, where are we now? We are the first ones certifying a Level 3 system. So obviously, we are leading -- absolutely leading the pack when it comes to autonomous driving, but even driver-assisted Level 2 system account Level 2 system of Mercedes plans for reading all this tests around the world. I think it's rated really highly. So I think we are in a good shape there, and we just talked about what is coming. In terms of disengagements, if you look at the disengagement report and if you look at the OEMs and disengagement report, probably there's only 1 on the top 10, which is an OEM. That's basically -- the other ones are basically Robotaxi companies that have a completely different business model. And there's a whole mix of data going into disengagement Formwork projects that we're running with Bosch, the Alsina project, which we discontinued and some future. So we are very happy actually with the actual performance of the system. And I would say we are on track there, as we said earlier, to achieve our targets. And what can you do in the car once you are in Level 3, it's about content and entertainment, immersive entertainment, and that's what we're going to see in the workshops there. And that's exactly why we focus so much on the canvas, on the light, on the sound on the massage, on the fragrance and everything that touches your senses in the car and make it a whole concept. And I mean, this world opens just up and we need partners like ZYNC. We need partners like Apple, and we're coming in new models with Dolby Music Universal. I mean, what we thought that we would come with these partners together in a project probably 5 years. We didn't think about it. Now we're tapping into a new pool of partners and thinking about new business models there. And that's just the beginning.

Thank you very much. So on that note, thank you very much for the very lively Q&A. Yes, and thanks a lot to everyone on stage for answering the questions. To all participants online, have a nice rest of the day, and we look forward to talking to you soon. Thank you, and goodbye. To all guests here in the room. There are still a few housekeeping things. We'd like to invite you in a moment to join us for lunch. I have just a few short remarks about the rest of the day. You can now continue your individual program over the afternoon. As you know, we'll be offering right along in Level 2, Level 3 ADAS vehicles. There is an active safety parkour and there's 3 different workshop stations to give you a precise picture of the tech innovations of the new E-Class. You've seen that during the presentations. The meeting point for the workshops is in the cafeteria that's over there, if you go out here and then over there. For the driving activities, please head to the Infodesk at the lobby. And if you do not know what your program is, please scan the QR codes on the backs of your name batches for your personalized schedules. Obviously, please feel free to explore the vehicle displayed outside as well. A fleet of Mercedes-Benz vehicles is available for you to choose from and drive around the building. To arrange for a car, please speak with the team at the test car counter. Our Board of Management members and the further experts will be available for discussions during this time. Shuttles for departures at the end of the day will be leaving from the front of the building over there. And if you will leave only tomorrow, please feel free to join us this evening for an informal dinner at 7 p.m. at the so-called EarthDoc restaurant. Let us know if you're in. So once again, thanks so much for your attendance, for your interest. Enjoy the rest of your day with your individual experience lots. As always, our team is at your disposal. Happy for any follow-ups. Thank you, and goodbye.