 Hello again. I'm Deb Bryant and I am grateful to be moderating and leading the discussion in our automotive panel today. In discussing with my fellow panelists, we think it's very important, especially because of the hour of the day, that we have a more engaging panel and we would like to invite the audience to participate. OFE staff have offered to bring a mic to anyone who might have a question. So we'll do some introductions, some rounds of opening comments. But if you have a question, please participate and bring those to us. We're actually all as interested in understanding what your curiosity is and what your ideas are as speaking ourselves. So following the comments this morning about needing to give better examples for how policy, and this is a policy summit, may impact the open source ecosystem. It's actually great to have a concrete example. And certainly the automotive industry, which is well centered in Europe, can serve as a great litmus test on how open source adoption is going to provide ideas on how it's innovating, to provide a state of the source and automotive, and also to maybe make some cautionary suggestions about how innovation may be hampered by example through our discussions. So I will start my name is Deb Bryant. I'm here representing Eclipse Foundation. And my perspective from the automotive industry comes from a number of years I was working for a commercial vendor and working with automotive companies on their emerging interests and concerns and participating in open source projects. I think I have the distinction of creating the first position for a dedicated community architect who was focused solely on upstream and community source projects and standards that have affected our automotive clients. So with that I'm going to ask each of the panelists to introduce themselves and also give you a little bit of context and where their comments come from their experience. So start with Martin. Yeah, great to be on stage with you. Yeah, I'm Martin Schleicher. I'm with Continental, responsible for software strategy in brief thinking about what software defined vehicle means for products and business models going forward. I'm also a member and representing Conti in some of the consortia like Eclipse software defined vehicle that has been mentioned here today. And as an open or additional note, just last week we published and first manifest or encouraging our employees to actively contribute to open source. That's fantastic. Thanks, Nishal. Hello. First of all, thanks for this opportunity to be part of this panel. Myself, Nishal Muchikani, responsible for strategic alliances, partnerships and developing the ecosystem for automotive vertical for Red Hat. And maybe first thing that most of you would identify Red Hat as a technology company more as a predominantly open source Linux cloud. So many of you may be wondering what we are doing in automotive. And I'm very glad that we have Carr as one of the team for a specialized panel discussion today. Thank you. Hello. Thank you very much. My name is Katarina Marakke. I'm a lawyer by training wearing multiple hats. And one of my first jobs was with Creative Commons. And on a private note, I was so happy and pleased that we also have a little Creative Commons reunion here in this community. But then followed a couple of other positions. And I have the honor to participate in this panel today in my capacity as a strategic advisor to ambition, which is a full subsidiary of Mercedes-Benz. So I think we have a good starting point for constructive discussion here around the car. Thank you. So you'll notice that we have representatives from different corners of the ecosystem. We don't have a manufacturer on stage, but we have a lot of insight and perspective from a fuller ecosystem. So I'd like to start by asking each of the panelists to give their perspective on what the state of deployment and adoption needs to open source is in the automotive industry from their perspective. And remember, we're talking about a range of areas. It could be entertainment systems. It could be operating systems. It could be self-driving vehicles. So we're interested in hearing what aspects of the automotive industry and their engagement open source. So I'll start. Let's go backwards and start with Katharina. Thanks, Deb. Yeah, I think when it comes to cars and the automotive industry, I'd like to share two observations. The first one is that the automotive industry came late to the party. So at a time when many other industries were already successfully working with and on open source, the automotive world and the OEMs in particular still thought that, you know, software in general and open source in particular is not relevant for them, right? Because they were focusing on hardware. And they were doing that perfectly well. So I think it became really only relevant when they realized that, you know, software is a big chunk in the value chain. And most of the big OEMs started to, you know, focus on in-house software development. And then open source became a topic for them. So we should keep in mind that everything we do with the automotive industry comes in this bigger context of the very difficult transformation from being a pure and very successful hardware company to actually a software company. And the second observation is that also traditionally collaboration in the automotive industry was very much focused on the supplier network. And that was very much tightly knitted around specific groups of suppliers, right? I think Martin will agree on this. So collaborating with external partners or perhaps even competitors was something that, you know, the automotive industry in general is not really used to. It doesn't come naturally to them. And it's paired with a fear of competition issues, right? So there has been this history of, you know, antitrust issues and that is still a red flag in the automotive industry. Now, when we keep that in mind and we remember what Carlo Piana said early this morning, open source is not an end in itself. Yeah, it's a means to an end. It's a tool for collaboration and for collaboration on predefined terms. So I think historically this way of collaborating with others on specific predefined terms is something that is not really organically or naturally grown in this particular industry. So keeping these two observations in mind, I think the learning curve has actually been quite nice, right? And we need to keep helping pushing and transforming that industry. Thanks. Thank you. Nishal, could you reflect what you're seeing with Red Hat's customers and your engagement in the auto industry? Sure. I think on the left hand side, we have a car maker on the left hand side, supplier and Red Hat as a technology provider. Most of you wouldn't realize what are we playing a role in the automotive industry? I mean, when we look the cars, as we call today as a software-defined vehicle of today and tomorrow, is, from a technology perspective, is an IoT device in the cloud. And when we talk an IoT device, the most complex ever built IoT device, because it's a convergence of so many technologies. You have the cloud piece, you have the big data every car, spitting out petabytes of data. Then you have the various connectivity piece that connects car to the outside world, 5G, Wi-Fi and whole host of things that leads to so many cybersecurity issues. Then we have the transition to the EV world, which brings in the e-commerce, the connecting to the smart grids and a whole lot of things evolving. So when we look as a car, it's a complete ecosystem outside connecting to the various technologies. And when you look all these technologies, open source is already there. It's a fundamental building block for any cloud ecosystem. If you look at telco, the transition, what we are telling happened a decade back. It was similarly probably 15 years back, the transformation has happened in the telco world. So all around the car, open source is already prevalent. And that's where we are already a major, what do we say, central offering, providing to telco industry, to the cloud and all those things. The most important of the specialized piece of technology is what goes inside the car. This is, I would say, the final frontier because this has a specialized ecosystem. This has a specialized piece of software where we have to be functionally safe. So when we talk about the car, the piece where we are tackling is trying to bring in, how can we bring in open source technology inside the car. So when we look at the end of the day or the ultimate goal is spanning everything from inside the car, way up to the cloud, what we call as the software defined vehicle ecosystem. If it is based on open source, I think that is where the industry can expedite its development. What we talk, the typical development life cycle of a car starts four years back. So what we are looking at in the current car is probably a technology which has been developed over four years. So when we build this kind of a life cycle and how do we expedite the automotive life cycle development, I think that can happen with the open source foundation and that's exactly our focus. As we always say, make OS boring or the infrastructure boring. It's more like you take it for granted it should work and you don't have to recycle the infrastructure platform to build and that would be our goal and that's the role what Red Hat and the community with which we are working is trying to build. Thank you for making a more boring product. It sounds safe. Martin. Yeah. Adding to the statement. So technical look at a car. So modern cars have like 20 up to 100 electronic control units of different types from a powerful infotainment system to very small dedicated electronic control units, for example, for braking system. That also shows that how far spread the adoption of open sources in the vehicle wind infotainment system, especially for connectivity and infotainment functions like MP3 playback or so open source has been used for more or less 10 years and you can look up the about screen in your infotainment system, what kind of open source is included in breaking and in other safety critical systems it's not used at all currently. So we'll have to look at the different areas of the cars individually to see where open source is already being used and where it could be used more software defined vehicle. This trend in seeing the and managing the value of the car through software and the complexity of software makes it more visible that the complexity can only be handled by collaboration, as Catherine said, and come industry, manufacturer, suppliers increasingly understand they have to collaborate no single company has the capabilities, the size and the skills needed to do it by themselves. So and that mindset change is currently happening. So and actually to also give you a practical example how the automotive industry has been working, how that's being transforming. So the industry has been working together in the past. So but usually on a specification level. So one of the hobbies I have is in a consortium called navigation data standard, creating a specification for a common map data format. That means in the past, architects, specialists from car manufacturers, suppliers, map companies have been sitting together in a series of working group meetings and creating a joint specification, which is usually the super set of all possible ideas. And then once the specification is approved, companies start building the software and the tools, and then maybe two years, three years later, we see the software and then we can start developing in the car. And I think this example shows that this way of collaboration is not the way to go forward because it just takes too long and the products are the hell complicated. And that's why we need to adopt in the automotive industry the concepts like code first, which has been mentioned by Gell, just in his keynote. And then start building something and see how we can evolve and join the collaborate on that and improving and collaborating in that in that NDS association. We have changed the mode of working in like, now we have an API that's that's open available. You don't need to become a member to use it. And that's abstracting the data which can be distributed somewhere in the car in the cloud and somewhere. And these are examples that the industry is now changing. And yeah, maybe we need a little bit more software defined management to have that also understand the sea levels. But I'm happy to see that industry is now moving. Then I'll put you on the spot a little bit more. So you've done a good job of expressing the state of the evolution of software development. What is the state of collaboration in the industry? Do you have most of industry coming along recognizing that this model is important and that individual manufacturers don't want to go on their own? They're not value? Are you at a tipping point? Or are you still a strategic advocate for collaboration? What's the state and the industry is it going well? Yes. Well, I'm not sure. When we'll see you really the complete source code of a certain system function being available and an open source basis using open standards for collaboration. I think that's the area where we are now and the industry to agree on how to define and leverage open standards to enable to build on top of what others are doing and create new innovative stuff. The competitive situation that Katrina mentioned initially it's still something that the automotive industry needs to learn that sharing of knowledge is actually helping you and not you're not necessarily giving away your competitive advantage. Thank you. Appreciate that. So I don't remind the audience that we're open for questions. I cannot see the audience so be sure and raise your hand and let one of the OFE staff members know and they'll bring you a mic. All right. Well, so let's let's talk about innovation a bit. So we kind of level set on what the state is and the industry engagement. What is who would like to take the question? How is software redefining the value of an automobile? What is the shift in the value of an automobile? I can start and then I'm curious to hear also your perspectives. So I grew up with this concept of open innovation, which has, you know, been the foundation to the whole concept of open source, right? So the way you innovate is also outside the traditional, you know, framework of a specific company organized by legal matters to, you know, keep straight secrets and knowledge in that company and open innovation was something that goes beyond that, right? So if we apply that to the today's state of innovation, I would say there's still work to do in the automotive industry, because as I said earlier, this way of collaborating with, you know, partners that have not been traditionally onboarded with traditional contracts and getting your developers to really be able to contribute. And I mean, I can I think this is not a secret. So ambition has released a contribution policy. So we are ready to do that. But it took a while, right? It was not easy to get that through the company to explain what it means for, you know, the own in-house developer to actually put something on GitHub. Yeah, so that is something that needs a lot of explanation and making the connection to what this means for innovation is even more difficult to explain in the traditional automotive industry. That being said, as I indicated in the beginning, I think we are at a good starting point, right? So the projects that have been mentioned hosted at the Eclipse Foundation with a code first approach, I think that is something really valuable and we are pushing hard for everyone to contribute to that. And if we meet again in one year, I was hoping that we could get some nice statistics on contributions and commits on that, right? Can you expand a bit, use the expression code first a couple of times for the audience that are not familiar? You would like to take that question. Can you repeat the question? To describe the code first? Okay, so when we look at the software defined vehicle, there are two life cycles we see. One, life cycle, which typically doesn't change during the course of the entire life cycle of the car, like the braking systems, the airbag controllers and all those kind of features. Then there is a second level of software life cycle, which is a consumer life cycle that the infotainment system, the assisted driving, the camera sensors, which changes almost, which needs a software upgradability features almost on a weekly daily basis through various software upgrades. So we need a kind of an infrastructure where you need a repeated upgradability. You need to have a software which is readily available for that to be upgraded. And that's where what we realize is to get a continuous streamed of certified software that was available in the open source world, especially when we look into the inside the vehicle kind of a constraints we have where it needs to be safety certified. We usually used to have a snapshot of a code that's available at a certain point of time, but if you need to have a vehicle constantly upgradable, that is where we are providing that open source community and where you have a code that's available for all the ecosystem to contribute and make it effectively available at any point of time. And for example, any change or line of code in a functional safety software would need a complete package of justification, documentation, everything to be available. It's not like you have a licensed version where you take a specific license and anybody can use it. And that would be the magic of the open source software where if it's available in public that can expedite a lot of this code development for the developer community. So we have about 10 minutes left in our panel and I'd like to drift into some adjacent areas that I think the automotive industry has a great skin in the game, so to speak, and that's the areas of data, open data and AI, because we it's kind of flipped to describe an automobile as a data center on wheels that's not very elegant, but there's a lot in coming up and going on. Maybe I'll start with Martin. We had a conversation about the importance of data and in the context of potential regulation I think it's good to have a public discussion. Yeah, I mean a lot of the software functions are more and more being based on AI, so simple example, traffic sign recognition, for example, speech recognition, that's AI based system. These systems need to be trained and if you think about traffic signs they may change their different and different geographies, so retraining the AI systems on a continuous basis is essential for keeping the performance of the vehicle in a good state. So access data and access to data is very important. On the one hand GDPR and the privacy and the rights of the end customer need to be protected, but also the access of data for the automotive industry must be granted. We have the EUA Data Act out there, which is taking care of some aspects. Here we as suppliers see one small issue because usually data, if the cars are sending data they are transferred to the car manufacturer's server. There's a secure connection and the car manufacturer has and gets access to the data, but we don't see adequate regulations also taking the roles of the suppliers into account also making non-discriminatory access to such data available. Our products are depending on that and we can only improve our products if we get a decent and fair access to the data. So maybe that's something also as a message to the regulators in the room here. Thank you, appreciate that. So the the safety of your product and the accuracy of your product may be based on what feedback and data you will get. Exactly, exactly. We take certain assumptions when we design the products. For example continental manufacturing brake systems, the better we understand how the data or the products are being used there and where, the better we can design our products going forward. Thank you. Katharine, do you have any insight into where you see the data concerns in such? Yeah, I think data is one of the most important aspects because the car is at some point just you know one item in the whole IoT system, right? It's connected with not only other cars and the environment but also most importantly with you know all your other personal devices. What I can share is I think five years ago I saw a presentation by someone from Tesla talking about their ADA system and their AI data set, right? So how many, how much data they have actually been able to gather and it was immense. I mean even at that time it was already clear that European car makers cannot easily or maybe can never catch up on that, right? Because they are just so far ahead of us just looking at the data. Now how did they manage just by sending the cars out on the street and doing things, right? We in Europe, I think we have a slightly different not only regulation but also awareness when it comes to data. So we, we are not as easy with just doing things and asking for permission later. I think we want to get things right from the beginning. So looking at that approach in GDPR is one example that's across all industries. I think the environment is, I wouldn't say more difficult, it's just different, right, from other jurisdictions. And coming back to the point of AI I agree that it's, it's really one of the important transformations in the future. What I'm a little bit worried about is the competitive advantage of Europe compared to other jurisdictions. At the same time, I grew up here, I'm very happy that we have this awareness and that I can opt out on many devices, especially also in the car, right? So I have the assurance that I'm not constantly monitor, I mean I'm monitored but I have the assurance that it will not be misused. So it's a, it's a two-fold issue, I would say. Thanks. Nishal, what about the perspective of IoT, AI and data? How is Red Hat thinking about that? Yeah, I think one of the unique features of the car is it spans a life cycle of 10, 15 years. And that's a generation which the consumer world couldn't comprehend. I mean if you look at phones, it's probably two years, three years. And here we are looking that the data has to be managed for 10 years. And probably it's not a problem of day one operation, it's a problem of what happens in the year number eight. Because you have millions of car on the road with different software versions and all this data has to be managed by the car maker through the entire life cycle. So we have a concept called a software factory that manages the versioning, the manages which car has what kind of data. And I think this is exactly where a completely new data modeling and data structuring will have to come by the car makers. And this would include ever evolving cybersecurity regulations because there is WP 29, the UN regulations that is evolving which is still in a very nascent phase. And if you really want to incorporate that over a period of 10 years, all the changes, your system has to be scalable and flexible to accommodate those kind of changes. And that's where I think basing a platform on open source will really help to adjust, accommodate and grow these platforms and also accommodate all the evolving changes. That's a technological problem we see from a technology industry because we provide technology for all the continents and we wouldn't do that any specifically for a region. That's why we have various partner programs called the ISV programs wherein the specialized solutions which come from the ecosystem, we certified to run on an open source technology platform and that is the kind of ecosystem how we have worked with other industries, banking, which is also a heavily regulated industry in a different sense of the world, telco and we have been able to show its works and it has been proven over decades. So automotive I think we will have to work as an ecosystem to develop that. Thank you. So we have three minutes so I'm going to divide it amongst each of you and put you on the spot. What do you each think is the we'll start with the thing that you worry about the most and the thing you're the most optimistic about? Who wants to go first now that I've just sprung a question on you want to start Katharina? So I hope to be an optimist by nature. So I think that through collaboration we will achieve a lot especially also you know on the topic of AI we need partners, partners outside the traditional supply networks. We need the code first approach not literally you know focused on software but to all other innovation mechanisms. Negatively speaking I still do see friction in terms of what regulators are trying to achieve and what the industry really needs and I think that dialogue sometimes I would say could be improved. Unfortunately I don't have any concrete. That's okay the parting thoughts. That's a party thing exactly. But I think you everyone knows what I mean. I think that dialogue is important to be improved and continued. Thanks. Thank you so much, Michelle. Yeah if you really look in last one year I think the ecosystem has collaborated. We have multiple initiatives in the industry. For example there is Sophie that is software open architecture for embedded edge there is Eclipse SDV which was earlier pointed out. There is Eliza embedded Linux for safety application. So which has all come understanding what's the industry problems. We have chip vendors car makers tier ones technology providers all coming together to develop a common framework. I think that is a very positive development because it's the first time I see the agility in the ecosystem to address such an issue. So that's a positive part from the ecosystem perspective. And we also if you look at automotive there is a lot of newcomers who have come into the ecosystem starting with Red Hat. And I think that will shape how the next generation automotive industry will evolve with fresh ideas and the new outlook of how to solve the problem. Fantastic. Thank you. Martin. Yeah on the positive side I think we're now really at a tipping point to see the way of open and open source and collaboration and automotive industry is really changing. And I would like to encourage engineers young engineers or female engineers in particular to contribute and be active. On concerns I actually wouldn't like to name three things. One is companies still are worried that speed and reacting to consumer demands and changes will compromise safety. So we need to get both done. And that's one of the topics. The openness is not in all organizations unfortunately visible across the management levels. So that's the other thing that we need to tackle and work upon. And the last thing is that open up and sharing information still is seen as a kind of giving away your crown jewels. And that's maybe also something that we need to tackle. Thank you so much. Let's thank our panelists for the session. I appreciate it. Thank you.