Name: Francis Chow | Red Hat
Uploaded: 2024-05-08T18:30:00.000Z

Clips
News
More from Red Hat Summit 2024

Revolutionizing the automobile industry: How AI and open-source software are driving innovation

Generative artificial intelligence and open-source software are used across various industries, and up until recently, one of the most often overlooked is the automobile industry, which has a large and innovative use case for the technologies. One of the biggest uses of AI and open-source technology in automobiles is the use of safety-critical systems in software-defined vehicles, an example of how these emerging innovations have applications across all market and are revolutionizing the automobile industry.Red Hat’s Francis Chow talks with theCUBE about revolutionizing the automobile industry

play_circle_outline Introduction to the Red Hat Summit and guests

play_circle_outline The concept of a software-defined vehicle

play_circle_outline The use of open source in the automotive industry

play_circle_outline Innovations in software development and deployment for vehicles

Info

Francis Chow | Red Hat

Rebecca Knight

>> Hi, welcome back to theCUBE's live coverage of Red Hat Summit here in Denver, Colorado. I'm your host, Rebecca Knight, along with my co-host, Paul Gillin. We are joined by Francis Chow. He's the VP and GM of In-vehicle Operating System and Edge at Red Hat. Thank you so much for coming on theCUBE.

>> Thank you for having me, Rebecca, and Paul.

Rebecca Knight

>> So before the cameras were rolling, we were talking about just all the excitement here that we're seeing and we're hearing about with medical breakthroughs and technological innovations. And we're here today to talk to you about autonomous vehicles. If you want to give us really high level of where we stand with autonomous vehicles and where we're at right now, we really had thought they'd be a reality by now. They're not.

>> Well, I think I'll probably step back a little bit. Instead of talking about autonomous vehicle, which is really, I think a facet of the whole picture, which we generally talk about software-defined vehicle. If you don't mind, we'll probably use that term for the rest of the interview. The industry has been talking about software-defined vehicle for probably a decade. And I think many steps from the traditional way of designing vehicle to the nirvana of, okay, through autonomous driving. And there're many milestones along the way. I think we are getting maybe about a third of the way. There's definitely a lot of technical challenges that we're working through with the ecosystem in the industry. There's always a balance of safety, security, and features, and time to market. And then there's also some more longer-term issues, including legal or ethics issues, that you get into for autonomous driving. So I would say the industry is definitely moving progress. Part of the challenge is that the traditional way of designing vehicles are no longer scalable and sustainable. And I think there's general consensus in the industry that open source could be an enabler for that transformation, and this is why we're here.

Paul Gillin

>> The vehicle industry is notoriously proprietary. Everyone does things their own way. How receptive have they been to this idea of using open source at a fundamental level like this?

>> I think there's definitely a transformation that the industry have to go through. I would say it's almost a necessity. If you think about how car has been designed for the past 40, 50 years, can you imagine in IT industry where every time you have a new model in a new production year, you almost have to redesign the software stack? But this is what the industry had been going through. So I think we're getting to a point that the demand for even driver-assisted systems, more personalized infotainment features, digital cockpit that give you more personalized driving experience these are getting to a point that the software complexity can no longer be managed by a single kind of closed-source render systems. And I think one of the ways to solve that is to allow the non-definitioning features of the stack to be open source, because whether using Linux distribution A or B doesn't really make a difference in a driving or driver's decisions on the purchase decisions, but these are the stuff that Red Hat, along with some other companies, can help the industry make it easier to design new software architectures.

Rebecca Knight

>> Can you talk through some of the innovations that you're working on now that will help transform the automotive industry?

>> Yes. So one of the... I think critical change the industry is going through is, how do they make software development and deployment more scalable. Because in the past, you design your car software. You get it to production. You forget about it, right? There's no need for updates and upgrades. In software-defined vehicle, we're talking about frequent updates and upgrades. It's almost like your phone, right? New apps going to be downloaded from time to time. The OS need to be updated from time to time. And so what we're trying to bring is the experience that we have with Enterprise, which we have been supporting mission-critical applications across multiple industries like healthcare, airlines, banks. They require as much quality, reliability, and security as the automotive industry does. So we're trying to bring that expertise over. Now, one key thing that we are working on is functional safety certification, because we do need to have that safety certification in order for us to be part of the vehicle technology stack. So this is the big piece that I think we bring to market, and it's really one unique differentiation that we think we can help with the market.

Paul Gillin

>> Maybe we can back up a step. What is a software-defined vehicle?

>> I think there're many definitions, as you can imagine, but my definition would be a software-defined vehicle is really a card that allows frequent update and upgrades of software and services. And the design methodology is no longer hardware-centric or waterfall-like approach that's in the past, but more like agile, like modern software system architecture and design methodologies.

Paul Gillin

>> Where do you see the innovation taking place? I mean, what do you see on the drawing board right now at automakers that makes you go, wow?

>> I think it's the thinking about how the car interact with the rest of the ecosystem. Because if you think a software-defined car, it's really not just about the car anymore, right? It's about how the car talks to this infrastructure, how does it talk to edge operating centers, how does it talk to the cloud because, in the end, there will be a lot of data that will probably be drawn from the vehicle and be analyzed in the cloud. There will be a lot of updates and upgrades that are going to be sent from the operating center to the vehicle. And I think what's exciting is how we can unify that design architecture so that the same applications can run in the car as in the cloud. Not a big surprise for IT markets, but it's a big change for the automotive market. How do you utilize the same tools to design all the functional domains in the car, which wasn't the case before? I think those principles that I think has been solved in IT for a while, then now we are trying to bring those into the industry.

Rebecca Knight

>> Well, before the cameras were rolling, we were talking about your career, and this is a relatively new area for you. And you described yourself as a changemaker, and this is what you specialize in as change management. How do you approach this? As you were just saying, it requires cultural change to think through these things and to be prepared for them. How do you approach the kinds of changes that are needed within the automotive industry and also within Red Hat too, to make some of these innovations realities?

>> It's definitely very exciting, personally. I think, as in many changes in technologies and in companies, we need to identify partners in ecosystems because no one can do it alone. So what we've been trying to do is to find like-minded companies and innovators in maybe more traditional companies and, "Hey, how can we work together to bring this to fruition?" And a lot of the things that Red Hat is pretty good at is to work with communities where there are a lot of automotive communities that we've been working with in the past three years to try to bring some of these concepts into fruition, including organizations like SOAFEE, Eclipse SDV, and ELISA for safety. So we're trying to share our knowledge and know-how with the community, take input and feedback, and then try to bring that together.

Paul Gillin

>> How do you think AI is changing the game, and what you do?

>> I think it's going to probably impact the industry in multiple ways. Similar to actually what we talked about in the keynote yesterday, there will be AI algorithms. In fact, some of that has been happening in the vehicle for the past couple years. When you talk about advanced driver assistance systems, there's AI in it because we need to make decisions like a human. We are seeing some even chatbot-type of features in certain vehicles, and it's actually pretty interesting because they are asking you as a driver what you want the car to do, and you can talk with it with natural languages and even provide feedback on what you want the product to look like in the future. So you got real-time update on feature requests from your customer base. And I think in terms of tooling, AI can also improve the design tools to make some of the software designs quicker and faster.

Rebecca Knight

>> When it comes to safety, do you anticipate that we will see an increase in safety? I mean, I'm just thinking about the harrowing statistics we're seeing about distracted driving, and my daughter is on the cusp of getting her driver's license very soon and having her on the road with these machines that are making decisions in ways that I don't know. Is it better to have a machine making a decision or a human? I mean, what do you foresee as sort of the future of roadway safety?

>> I think there will probably be a transition, right? In the near future, I would say probably the next five to 10 years. I would say human-assisted driving is still required. And in the same time, I think the industry... In fact, there's some forums and communities, for example, we are part of a World Economic Forum initiative to talk about these problems. And one of the things that we talked about recently was it's actually a challenge because every car makers have a different human interface for assisted driving. Now, think about if you own a certain car brand and you change your car-

Rebecca Knight

>> Or you rent a car, or-...

>> it's -

Paul Gillin

>> It's not like a computer-...

>> experience-...

Paul Gillin

>> right? It's like a GUI.

Rebecca Knight

>> Right, right. Yeah.

>> So how do we make it easier because that's part of safety? It's not just technology, but it's how the vehicle interacts with humans.

Paul Gillin

>> We'd be remiss if we didn't mention that your title also includes edge computing, which is a great many more use cases. What are some of the more interesting edge computing field applications you're working on right now?

>> Looking at all edge broadly, because we think similar to automotive, what we're seeing is there are multiple industries that are getting to a point where they require a transformation, and they all sort of agree that open source would be one of the enablers of those transformations. So we've been investing heavily in manufacturing, retail, energy infrastructure because those are the industry that we see transformation is at the pivotal point of change. So we are working with our partners in ecosystems to build solutions that would fit the needs. And I think some of the changes that we seeing is there are definitely a change of decades-old model. For example, for manufacturing, there's a well-known model called the Purdue model. How do you segregate a control system into different layers of control, from a device level to your daily acquisitions to your execution systems? Some of that is probably going to collapse and change so that they can take advantage of new IT concepts. And some of the cool things that we've been talking about and working with... I'll give you one example. We partnering with a company called Minsait in Europe, and we implementing a solution where they can use AI. We talk about AI all day. So that the wind farms can detect flying birds. They can change the speed of that turbine to avoid collisions of the birds.

Rebecca Knight

>> Wow.

>> Yes.

Rebecca Knight

>> Right.

>> Isn't that cool?

Rebecca Knight

>> That's very cool.

Paul Gillin

>> Illuminating one of Donald Trump's major objections.

Rebecca Knight

>> Exactly. You won't have any more complaints. Final thoughts? What... I mean, this conversation has been so illuminating and fascinating. What do you think is on tap for the year ahead, and what are going to be the top conversation topics of next year's Red Hat Summit? As you said, AI is obviously the buzzword this year, and it will continue to be next year too, but what is on the horizon that you're most excited about?

>> I think the intersection of AI and the Edge would probably be... It might not be next year, but it'd probably be down the road, be a very exciting area. Because in the end, if you think about what's most probable of leveraging AI, is where the actions happens. The action happens at the edge, not at the cloud, where the AI models are being trained. And if you think about where the data is, the data is at the edge. So why not training the model at the edge, leveraging the data at the edge, and then deploy it where the model and the data is generated? And I think with advancements in shrinking of model sizes, with advancement in increasing of compute power of small devices, with advancements of other factors to technologies, I think we going to see a boom in AI inferencing at the edge.

Paul Gillin

>> So I'll put you on the spot and ask you for a prediction. Will a child born today have to learn how to drive?

Rebecca Knight

>> Good question.

>> Very good question. As a parent, I think it's a good skills to have, just like swimming, but it might not be a necessity. And it could become a leisure thing, just like I would love to drive because it's fun to do.

Rebecca Knight

>> Right. Great. It'll be a hobby for our grandchildren. Excellent. Well, Francis, this has been really fun. Thank you so much for coming on theCUBE.

>> Thank you, Rebecca. Thank you, Paul.

Rebecca Knight

>> I'm Rebecca Knight for Paul Gillin. Stay tune for more of theCUBE's coverage of the Red Hat Summit. You're watching theCUBE, the leader in technology enterprise coverage.

Francis Chow | Red Hat

Rebecca Knight

>> Thank you for having me, Rebecca, and Paul.

Rebecca Knight

Paul Gillin

>> The vehicle industry is notoriously proprietary. Everyone does things their own way. How receptive have they been to this idea of using open source at a fundamental level like this?

Rebecca Knight

>> Can you talk through some of the innovations that you're working on now that will help transform the automotive industry?

Paul Gillin

>> Maybe we can back up a step. What is a software-defined vehicle?

Paul Gillin

>> Where do you see the innovation taking place? I mean, what do you see on the drawing board right now at automakers that makes you go, wow?

Rebecca Knight

Paul Gillin

>> How do you think AI is changing the game, and what you do?

Rebecca Knight

>> Or you rent a car, or-...

>> it's -

Paul Gillin

>> It's not like a computer-...

>> experience-...

Paul Gillin

>> right? It's like a GUI.

Rebecca Knight

>> Right, right. Yeah.

>> So how do we make it easier because that's part of safety? It's not just technology, but it's how the vehicle interacts with humans.

Paul Gillin

Rebecca Knight

>> Wow.

>> Yes.

Rebecca Knight

>> Right.

>> Isn't that cool?

Rebecca Knight

>> That's very cool.

Paul Gillin

>> Illuminating one of Donald Trump's major objections.

Rebecca Knight

Paul Gillin

>> So I'll put you on the spot and ask you for a prediction. Will a child born today have to learn how to drive?

Rebecca Knight

>> Good question.

Rebecca Knight

>> Right. Great. It'll be a hobby for our grandchildren. Excellent. Well, Francis, this has been really fun. Thank you so much for coming on theCUBE.

>> Thank you, Rebecca. Thank you, Paul.

Rebecca Knight

>> I'm Rebecca Knight for Paul Gillin. Stay tune for more of theCUBE's coverage of the Red Hat Summit. You're watching theCUBE, the leader in technology enterprise coverage.