Clips
News
More from Red Hat Summit 2024

Stavros Stefanis

Principal, US Software Defined Vehicles Leader

Deloitte

Dan Noal

Director of Product Management

Red Hat

Realizing the future of autonomous driving: Red Hat and Deloitte weigh in

When Tesla Inc. began building out self-driving technology for its electric cars, the industry was thrown into a frenzy from which several talking points emerged — with excitement on one side and pessimism on the other. Since then, however, progress in that regard has been slower than expected. Red Hat Inc. aims to expedite the tech for autonomous driving with its radical technology stack.Talking about the infusion of open source in autonomous driving tech, with Deloitte’s Stavros Stefanis.“Red Hat believes that now is the time for open source and automotive,” Dan Noal (pictured, right), director of product management for in-vehicle OS at Red Hat

play_circle_outline Collaboration between Deloitte and Red Hat to transform the automotive industry

play_circle_outline Red Hat's In-Vehicle Operating System and its significance in the industry

play_circle_outline Impact of AI and advancements in technology on the development plans

16:24

15:44

16:02

Info

Stavros Stefanis & Dan Noal | Red Hat

Stavros Stefanis

Principal, US Software Defined Vehicles Leader Deloitte

Dan Noal

Director of Product Management Red Hat

Rebecca Knight

>> Good afternoon everyone and welcome back to theCUBE's live coverage of the Red Hat Summit here in beautiful Denver, Colorado. I'm your host, Rebecca Knight, alongside my co-host and analyst, Paul Gillin. Paul, we're talking today so much about the... Often on the show, we talk about the future of work. For the second time, today we're going to be talking about the future of driving.

Paul Gillin

>> Which I think is such an exciting topic because there are so many possibilities, but getting there is the hard part and I think we've all seen that. The move to autonomous vehicles has gone far more slowly than was anticipated six or seven years ago. Red Hat is going to try to accelerate it with their own technology stack.

Rebecca Knight

>> Exactly. And so who better to talk about it than our next two guests? We have Stavros Stefanis. He is the principal at US Software Divine Vehicles at Deloitte. Welcome, Stavros. And Dan Noal. He is the director of product management for In-Vehicle Operating System at Red Hat. Thank you both so much for coming on theCUBE.

Stavros Stefanis

>> Thank you.

Rebecca Knight

>> So as I was saying before, we had your colleague on... Your boss, Francis Chow on, so we've talked a little bit about these topics, but I wonder, Stavros, if you can give our viewers a high level overview of the collaboration between Deloitte and Red Hat that is aiming to transform the automotive industry.

Stavros Stefanis

>> Sure, Rebecca. So the industry is at a very exciting point. We went from the world of steel and hardware to we're going to the world of software. And we're working very closely, Dan and the Red Hat team with our team at Deloitte in a space called Software Defined Vehicles, which is all about how to make a vehicle product that's really the defined, experienced through software, developed as well through software. And the software doesn't pertain to just the vehicle itself, but also the broader ecosystem around it. It's a transformative space because in parallel with this evolution towards software, there is evolution in compute capacity that makes really the electrical and electronic environment of the car so much more powerful that you can run a lot more complex software and deliver a lot more advanced features with the end product. It also allows... If you think about it, the vehicle used to be a one-time sale. You would go buy a car, leave the dealership, and then two, three years, maybe five years if you liked it a lot, you would go and trade it, get another car and so on. You would service it. Well, now we're moving to an age where the old car that you buy once you leave the dealer can be upgraded. No different than a phone, right? You can add applications, you can add features, you could even repair it sometimes while it's parked at your home with over the air updates. Very, very different world. So the car has increasing potentially monetary value post initial sale.

Paul Gillin

>> Very exciting. Tell us about the Red Hat In-Vehicle Operating System. What exactly is that?

Dan Noal

>> Red Hat believes that now is the time for open source and automotive. So we've seen industries transform primarily in the data center enterprise applications where Red Hat Enterprise Linux is obviously a leading commercial solution. And edge markets and specifically consumer automotive we think is the next opportunity. Today, the automotive industry software, let's say landscape consists from the notion of operating system suppliers is fragmented and largely proprietary. And we think that the advantages of open source really are paramount and foundational and necessary to achieve the vision that Stavros just talked about. That if we want to create innovation at pace and at scale cloud native development, really the transformation of how an industry develops software foundationally, that needs to be based on open source, to speed the pace of innovation to provide that standard foundation everyone can build upon. And that's what our partnership with Deloitte is all about.

Paul Gillin

>> But the auto industry is one of the most proprietary industries on earth. And I know you are working with GM on this project. How receptive in ear have you gotten from the others?

Dan Noal

>> Very receptive. The big issue is that Linux already exists in automotive today for non-safety critical systems like your infotainment system, like maybe communication gateways, there's other examples. But most of the, let's say software content in a car relates to a system that's safety critical and we have regulations and industry norms around safety and how software can be qualified to run in a functionally safe environment, in a safety critical environment. And it's been difficult up to now for Linux to meet the standard and attain that certification. Those certifications were developed with proprietary software in mind. Small, not complex. Typically, real-time operating systems with proprietary development environments. The world has come a long way. Compute has come a long way and the time we believe is now for Linux to be fully qualified for functional safety and for safety critical workloads. That's what we're working on and that's what the industry really is waiting for a full embrace of what we're doing.

Rebecca Knight

>> Well exactly, but embracing this concept of a software-defined approach, how are you positioning it? I mean, Dan just said that automakers are very receptive to this idea, but how do you bring more of them on board?

Stavros Stefanis

>> It's a transformation. So to what Dan is saying, we are at the point where the current way software is developed by the automotive industry, whether you are an OEM or a supplier or even a tier two company is not easily scalable. It doesn't scale. And the reason that it doesn't scale is because you need to evolve into an environment that's open enough to attract software engineering talent and software engineering capacity since most of the content of the car is becoming more and more software centric. So you need to be able to scale your software talent and the best way to scale software talent is to work in an open source environment. Now that requires a very transformative approach for the main reason is that specific domains are perceived legacy domains, that they need to get modernized. There's also collaboration that's required between different teams because traditionally people who think about infotainment in a car and they develop the software code for infotainment, they work in their own area. People who think about driving experience, they work in their own area. Once you start moving to a cloud native more open environment, then what you can do is you can containerize these different areas, especially as solutions get more and more safety certified, such as in this case, RIVOS gets the safety certifications, you can do containerizations and easily integrate the pieces together. Now that's not only beneficial from a software engineering perspective because then you can set up teams that they can seamlessly collaborate and move in developing their code very, very quickly in an agile environment, it is also very important because you can launch complex features. And I'll give you an example. We heard one this morning. I really like that. And the one we heard this morning is imagine a feature of a car approaching the car wash and the car itself talks to the car wash. You have no more these rails that scratch your tires. It happens automatically. The car shifts into a drive mode that is dictated by the car wash itself. So it self-drives through the car wash, mirrors automatically close. And by the way, you can do all these things from your phone. And put your car seat outside at the coffee shop, watch your car go in and there you go, you have it. Your car is not going to get scratched. No one even touches it.

Paul Gillin

>> Such a wonderful area for speculation is the Red Hat In-Vehicle Operating System intended to cover all aspects of operations. I mean you have navigation, you have monitoring of diagnostics, you have entertainment. Any number of other potentially software defined operations? Do you cover all of those areas?

Dan Noal

>> The answer is not all areas, but we think over time it's going to be most areas. The highest safety level for the car, which would be systems like your anti-lock brakes, maybe forward emergency braking, real safety critical features, have a higher designation than we're pursuing with our strategy around functional safety certification. And so those workloads we believe will still use legacy runtime environments and not an open source Linux that can achieve that highest safety level rating. And I think that's okay because most systems can be decomposed where at least some of their functionality can be implemented on an open source baseline. In these new style high performance computing, they're sometimes called zonal or domain controllers. These are new terms the automotive industry is using for the type of computers that go into them, which is very different from what you heard in the past. You heard there's an infotainment unit, there's an anti-lock braking unit, there's an engine control unit. Now they're getting merged and basically combined into controllers. And so we can isolate the highest safety level that those things where Linux doesn't apply effectively in that type of environment.

Rebecca Knight

>> I want to return to something that Stavros was talking about, this transformation that's taking place not only with the technology but in people's jobs. Because as you said, jobs were much more functional. If you worked infotainment, that's what you did. If you worked in drive, that's what you did. But now we're seeing many more people working cross-disciplinary and having to think differently about how they produce and how they collaborate and how they work with colleagues. I'm wondering how you are helping customers along the way with that kind of culture change. Because it really requires, and as you said before, the different kinds of skills that people need too, a real upskilling, a real culture of learning and upskilling. How you're working with customers to instill that too.

Stavros Stefanis

>> I'll give you a couple of examples. So one area that's critical in this is because you can develop now software almost agnostic of the hardware, and then as you go along and you develop the software, you can make hardware decisions in parallel, you need to be able to represent logically what you're developing. So that whole space is called systems engineering. That's an area where a lot of the automotive OEMs and across the value chain, automotive value chain, there is a big need to develop skills. And we're helping companies develop skills, processes, understand roles and responsibilities of those engineers and how they need to integrate in the organization, how they need to be measured. That's a very important area. That's one example. A second example and a big change that is happening is not only you need in this new era, really good software engineers, you need your software engineers to be close to the customer and understand what the customer experiences, understand what are the preferred features that are out there, understand what are the needs in terms of doing things quickly and at what speed. In the old world before, software engineers would prioritize based on hardware release schedules and they would say, "Okay, we have six, seven months, we develop these features. Hardware will be ready by then and then we launch." No more, no more. Now when there is a need to launch new things or even more importantly fix things, the window is very compressed. So the software engineer needs to be very in tune with the customer and what the customer needs are to prioritize accordingly. So that's another mindset shift that's happening. Getting software engineers very connected with issue management, with customer management, with claims, complaints handling, all these areas.

Paul Gillin

>> The Red Hat In-Vehicle operating system was announced three years ago before the surge of interest in AI and all of the technology that we've seen in the last 18 months. How is that affecting your development plans?

Dan Noal

>> Well, it's certainly having an impact, isn't it? These are the exciting transformation of entire industries. So interestingly enough, we always thought about portions of AI because you think of autonomous driving as one of the... or driver assist, those ADAS type features, are a big target of what we're trying to do with... And a big target workload we're trying to get to run on the Red Hat In-Vehicle Operating System. And there's always elements of AI in terms of inference and pattern recognition from different sensors. And that's only going to be accelerating with the advances in AI that we're seeing. And it's going to enable new use cases, faster time to market, faster introduction of improvements, all supported by this whole software-defined vehicle lifecycle that we're talking about.

Rebecca Knight

>> Well, I know that Paul asked a really great question of Francis Chow from our last one, which is, will a child born today need a driver's license? Will that be a life skill that a human born today will need? I'd love you to just think broad brush about where we're going to be in the future because we know that this has taken a lot longer, this push toward autonomous vehicles. When will we be there? As a closing thought, when do you think, Stavros, and I'd love to hear your thoughts too, Dan?

Stavros Stefanis

>> Dan, I'll go first. I think you have to see the mobility world in two areas. The commercial and the private, the more passenger-oriented. I think from a perspective of autonomous vehicles, the commercial space is moving and will move faster. Mainly because it's more predictable because the track goes from point A to point B. There's fleets, you can manage fleets. You can set up missions. So that space, the fleet management space, the commercial space, you're going to see a big, big change, the next 5 to 10 years. The same with local mobility within cities where the distances are smaller, you'll see the robot axis fleets a lot more in the future. I think that will happen in the coming years. Now on the other hand, in the more personalized space, I think you're going to see a shift to people using more the new means of transportation, robot fleets and so on. So relying more on these types of transportation versus the traditional ownership model of a vehicle. And I do believe eventually you'll have less licenses as well. Because let's face it, it's convenience. Who doesn't want to go on a ride and be able to work or maybe take care of their kid while they're traveling together, while they go from point A to point B versus worrying about traffic stop and go, et cetera?

Rebecca Knight

>> Me. I'll sign up.

Stavros Stefanis

>> There you go. There is always all the aficionados. I always like my drive in the countryside. So you'll have that too. So that's my two cents.

Rebecca Knight

>> How about you, Dan?

Dan Noal

>> Well, I think it's such a great question and I have two daughters that live in New York City now and neither of them drive in New York City at all. And it's not based on autonomous driving, it's based on convenient access to public transportation. And if I scale forward 15 or 16 years, think about being born now, that's only going to increase in terms of convenience. And they have to take occasional trips, into Pennsylvania or something where they might rent a car, and that might be an occasion where they would use an autonomous vehicle. And I think in 15 years, most of those cases are going to be handled in some fashion other than needing to have your own vehicle or drive your own car, even if it's a rental car. But I think in rural areas, you think of a country like ours with wide open spaces, there's going to be a lot of our country that won't be covered economically by Robotaxis and other automated solutions or public transportation. I think that's going to take a long time. I can't predict what's going to happen in 15 or 16 years. That's too far out. And we talked about earlier, there's an entire class of people in this country which represent huge percentage, they're going to want to keep driving as a hobby, as recreation. And the question is how long will that last. And I see it to be kind of indefinite, but we'll see.

Rebecca Knight

>> You don't know. Exactly. Well, this is really fun questions to ponder. Dan and Stavros, thank you both so much for coming on theCUBE.

Dan Noal

>> Our pleasure.

Stavros Stefanis

>> Our pleasure.

Rebecca Knight

>> I'm Rebecca Knight. For Paul Gillin, stay tuned for more of theCUBE's live coverage of the Red Hat Summit. You are watching theCUBE, the leader in tech enterprise coverage.

Stavros Stefanis & Dan Noal | Red Hat

Rebecca Knight

Paul Gillin

Rebecca Knight

Stavros Stefanis

>> Thank you.

Rebecca Knight

Stavros Stefanis

Paul Gillin

>> Very exciting. Tell us about the Red Hat In-Vehicle Operating System. What exactly is that?

Dan Noal

Paul Gillin

>> But the auto industry is one of the most proprietary industries on earth. And I know you are working with GM on this project. How receptive in ear have you gotten from the others?

Dan Noal

Rebecca Knight

Stavros Stefanis

Paul Gillin

Dan Noal

Rebecca Knight

Stavros Stefanis

Paul Gillin

Dan Noal

Rebecca Knight

Stavros Stefanis

Rebecca Knight

>> Me. I'll sign up.

Stavros Stefanis

>> There you go. There is always all the aficionados. I always like my drive in the countryside. So you'll have that too. So that's my two cents.

Rebecca Knight

>> How about you, Dan?

Dan Noal

Rebecca Knight

>> You don't know. Exactly. Well, this is really fun questions to ponder. Dan and Stavros, thank you both so much for coming on theCUBE.

Dan Noal

>> Our pleasure.

Stavros Stefanis

>> Our pleasure.

Rebecca Knight

>> I'm Rebecca Knight. For Paul Gillin, stay tuned for more of theCUBE's live coverage of the Red Hat Summit. You are watching theCUBE, the leader in tech enterprise coverage.