Autonomous cars as a service: How Panasonic is preparing for a future where most drivers don't own cars

Panasonic is preparing for a future where autonomous cars upend the auto industry and most cars are sold to ride-sharing companies like Uber or Lyft.

Autonomous cars as a service: How Panasonic is preparing for a future where most drivers don't own cars Panasonic is preparing for a future where autonomous cars upend the auto industry and most cars are sold to ride-sharing companies like Uber or Lyft.

TechRepublic Managing Editor Bill Detwiler spoke with Panasonic's Andrew Poliak about how Panasonic is preparing for a future where autonomous cars upend the auto industry and most cars are sold to ride-sharing companies like Uber or Lyft. The following is an edited transcript of the interview.

Andrew Poliak: We see this trend happening in the vehicle space where OEMs are rushing to autonomous and assisted driving technologies, and they're running out of physical packaging space. When you start thinking of all the microcontrollers in a vehicle, there's cameras, there's sensors, there's RADAR and LIDAR and all these things coming into the vehicle that they just don't have the space for wiring or components.

What we've been doing is taking our Spider platform and up integrating more pieces. The surround view camera which used to be an external module is integrated into it. The control for the cluster and the IVI, and the headup displays, integrated into the one system. The active noise cancellation for reducing unwanted noises out of the cabin, that's up integrated into the system. So we've calculated, we're saving almost $200, actually over $200 per vehicle of electronic costs by integrating that into one powerful system.

Bill Detwiler: Talk about that, the important of cost, because when you're producing tens of thousands, hundreds of thousands of unit, every hundred dollars that you save matters, every bit of space that you save matters. And it also matters when it comes to, as vehicles become more expensive, autonomy is supposed to help eliminate accidents. But when accidents do occur, the vehicles are more expensive to repair, parts are more expensive to replace. Can you talk a little bit about that in terms of what... As we pack more and more technology into smaller and smaller spaces, how do you mitigate that kind of... The cost of having to replace or repair, or even just originally purchasing the vehicle?

Andrew Poliak: I mean, one thing, it's obviously a lot cheaper when those cars are not crashing. Another thing that's happening is the electrification of the vehicle. That's reducing the typically wear and tear of parts. So, by going full electrical vehicle, and not having an internal combustion engine, that's actually a much less wear and tear, and much easier to build up a vehicle and even replace components. So you think about why it's easier for startup companies to create electric vehicle companies, or electric vehicles. I had one cell phone manufacturer once say, "An electric vehicle just looks like a big phone."

SEE: How 5G will transform business (ZDNet special report) | Download the report as a PDF (TechRepublic)

The domain knowledge you needed to have around the internal combustion motors and all that goes with it, that was also a very big expense, and capital expense. Replacing some of these things actually may be a little bit easier as you go and reduce the cost.

But you said hundreds of dollars. As I mentioned, we can save that much. We have OEM's that literally will choose a supplier that may not be as competent as another one over a penny, or over a nickel. So you know when you start saving that kind of money, they're definitely looking, and looking at you to say, okay, you can help understand how to drive the pieces to give us that investment to go after new technologies rather than things like entertainment, which are differentiators for the human experience, but are becoming commoditized. So how do we make sure that we can supply things to OEM's, make it cost effective, and make it able to scale, and make it to be a great experience?

SEE: 5G technology: A business leader's guide (Tech Pro Research)

Bill Detwiler: So you were talking about commoditization in terms of entertainment, behaviors, things like that. Talk about, are we at the point of commoditization with the sensor technology? Or is that still further down the road?

Andrew Poliak: Now, well, even in the vehicle entertainment side, I think there's going to be a new resurgence of branded audio and branded experience. That may change that again as you get to full autonomous vehicles, because you're suddenly an hour in your next best living room, you know? So we've had a number we've launched with Fender, with Nissan and the Titan. We've launched with other OEM's with branded audio, like Acura. We've got one, I think the audio system of the year with the ELS, Elliot Scheiner's ELS Studio.

So don't get me wrong, I think there is a lot of value in the entertainment going forward. But in the core platform, especially as Android becomes more popular, it's going to be about speed, it's going to be about servicing that system, and it is going to be about driving the cost out of those kind of commoditized components.

In the sensor area, I think it'll be a while before it gets commoditized. However, we try to take really expensive sensors, and how to replace them with things that are at scale and can be cost effective. For example, Panasonic is legendary for cameras. So, using cameras to do new sensors, so be able to sense objects, or do things that you would usually have to do with very expensive LIDAR, being able to do that with more easily mass-produced component is one way to reduce those costs in those autonomous and ADAS vehicles.

Bill Detwiler: Talk a little bit about the disruption that autonomous vehicles are bringing to the automotive industry that we were talking about earlier.

Andrew Poliak: Yeah. It's amazing, because it's not just autonomy, but also these ride share and service type systems. You know, every vehicle manufacturer, so the Porsche's CEO said, "We'd rather sell a Porsche 1,000 times than 1,000 Porsches." That mentality is starting to sink in that everyone's investing, GM investing in Lyft and Cruise. DD was invested in, I think, by Daimler. You name across the spectrum, everyone's getting into ride share and fleet, and also autonomous.

What's really being disrupted is, an OEM would customize a vehicle almost completely for a 30,000 taxi cab purchase in New York. So imagine when you have Waymo buying 62,000 Chrysler mini-vans and 20,000 JLR vehicles. We're seeing this disruption where our... The customer of most of our customers may no longer be the consumer, but may be a fleet, and that changes the dynamics of the supply chain.

If you look at what we're engaged with, a lot of these mobility companies, because they may specify to the OEMs, "This is the infotainment system we want. This is the sensor we want." And you'll see us engage a lot more with those mobility services as they are working on cutting-edge technologies that they may directly source to their manufacturers.

Bill Detwiler: Do you have any idea, or if you could get out your crystal ball and think about any move to a world where individual car ownership is not the norm, but it may be... it's the outlier? How far away? I mean, do you think we are there? Is it a generation? Is it much faster than a generation, much more quickly, because the pace of change is so much quicker now than it was in the past?

Andrew Poliak: Well, you know, it's funny. We have co-op students every year, and every year, they have an end of their co-op period, where they do the projects they work on and present to our executive team, "Here's the things we learned. Here's what we worked on." And every year I ask them, "How many of you own a car?" and every year, those hands get smaller and smaller as it goes down.

I think you'll actually find that digital divide of the digital natives versus digital immigrants is about that line, that it seems like if they're also in a major urban or metropolitan area, I think it won't even take a generation. I think you're already seeing people be less interested in getting driver's licenses. You're seeing people that don't have cars in the city already.

We did a number of announcements too. I think you'll find different forms of mobility. Think about all the Lime scooters, and things you see all around major cities. That's because there is less vehicles. So, we did our announcement with Kent and with Van Dessel bicycles that were bringing our Panasonic, 40 years of perfecting E-bikes in Japan, bringing that to the US with motors and batteries. We're working on two-wheel mobility with Harley-Davidson with the LiveWire, and the connected LiveWire experience. And you'll just start seeing these micro-mobility devices come out, and I think it will really accelerate how fast people can do without a traditional vehicle.

Bill Detwiler: Let's switch gears a little bit and talk about connectivity.

In order for these, whether it's autonomy or the new technology that people expect from cars, that requires connectivity, even more so than that. So, talk a little bit about maybe 5G.

How much of a game changer is 5G, or will be 5G, when we get mass adoption to transportation? Talk about it.

Andrew Poliak: I think quite a few things will become a big game changer from that. Just the amount of experiences and the data you can pack into it, and also how you can... Some of the 5G implementation going forward will have not just normal data communication, but also have cellular V2X integration in one common modem. That allows for peer-to-peer communication between vehicles and vehicles in infrastructure, as well as normal network connectivity. So, that's going to open up safety use cases, experiential use cases, streaming media and audio, diagnostics, you name it, all the fleet and ride share services.

We're already demonstrating here that we can run Netflix in protected mode in our infotainment system, and we do that because... When I say protected with DRM, or it can download content and store it locally with your Premium Netflix account or Spotify. We do that because connectivity isn't ubiquitous right now. It's not necessarily high-speed to the car, and people don't want to pay always for another connection with today's experience. But going forward, everything's going to be connected.

In the topic of connection, it's not just cellular or 5G, but every device you see in the car will have a reason to have a connection. Let me give you a good example. We worked on lane-precise GPS receivers. When you get to that millimeter precise kind of precision in GPS, you suddenly need to start correcting for atmospheric conditions in the communication of the satellites. So, you have to have a connection to get the service to do the data and the error correction for that precision. When you get into autonomous vehicles, even little sensors will have reasons for connectivity, and 5G will allow all those different sensors to have that kind of ability to update for precision, for correction, you name it.

SEE: What is 5G? Everything you need to know about the new wireless revolution (ZDNet)

Bill Detwiler: With all that connection, with all the technology always comes a risk.

Talk a little bit about security. How can we assure that the connections, that the devices don't open security risks in those vehicles?

Andrew Poliak: Yeah. I always go back to the story that was a few years back, and I know Volvo, I think, originated it, or Sweden, with a goal zero, about zero fatalities? But I used to live in Washington state, and they had a goal zero, and it was a great video they did. They went around to all these people and said, "What do you think the fatality rates in car accidents are a year in our state?" And people gave every answer for hundreds to tens of thousands of people. They had no idea, right? And then they said, "Well, what should our goal be?" And again, they gave ranges of hundreds, ones to thousands of people.

Then they said, "Oh, great. Now, what should the goal for fatalities related to automobiles in your family be? What should your goal be?" And everyone just looked like, and went, "Oh, zero, of course." And they said, "Well, shouldn't that be our goal for everyone?" That's the approach we're taking at Panasonic is you can't allow that first security risk, especially in an autonomous vehicle, because the trust you lose will set that industry back so far. Even if you could show that the overall value outweighs the risk, we're living in a world that it's obvious that small risk can be blown up to huge problems, and it can set back industries. It can cost our society a lot of money and a lot of time.

Just looking at it and securing the vehicle bus, performing these air gaps between safety critical components and non-safety critical components. What we're doing with hypervisor technology that allows you to run multiple operating systems and most multiple safety and non-safety environments together, but very isolated. Just really taking it like we're members of the Auto-ISAC, where we share security threats and risks with other suppliers and OEMs, transparently, quickly. That way, a lot of us use similar technologies, whether it's Bluetooth stacks or Wi-Fi components, making sure that we all collaborate and never allow that first really big problem occur.

Bill Detwiler: The last question I have, as cars become more technologically advanced, do you think that they will take the route of other electronics when it comes to updates? You know, since we're talking about security, we all live with security updates on a daily basis for our phones, for our computers. We haven't got there yet for the toaster or for the fridge, but we're sort of there.

But when it comes to the auto industry, updates were always an upsell. They were always an upcharge. It requires you to take the car in to the dealer. It was something that the manufacturers weren't that interested in providing those either for free, or... which is the expectation of the software industry. Talk a little bit about that, in terms of the automotive industry. How do you see the update process? You know, Tesla charges five grand if you want to get Ludicrous mode. Well, why am I paying you 5,000 for a software update, right? It's your code. I like your code, but I don't know if it's that good or not. Talk a little bit about that. How do you see that playing out?

Andrew Poliak: Yeah, absolutely. We're already in production programs now that can do over-the-air software updates. It's deploying to cars today.

What we'll see is there's three different kind of update solutions. There's what's called firmware over-the-air update. That would be for your low-level operating system, let's say, or other components. There's software over-the-air update downloading new apps, or navigation features, or all that. Then there's ECU Reflashes. That's downloading and then updating other little microcontrollers like Tesla does with changing ride heights or ride styles, and things like that.

Almost every program we have has those requirements, at least those three type of updates, along with... It used to be you'd build to a spec and deliver it. Now that specification is requiring security patches automatically planned into... You don't ever just ship it and it's done. You're always constantly working with the OEM and maintaining it over a period of time. Then in the future, there'll still be security updates that you'll have to include as well.

Going back to security for a second, it's not just the updates and the patches and all that. It's also anomaly detection, saying, "Hey, something's in here that doesn't seem like it is behaving the way we predict it should, so what is this?" So, using machine learning. That's also where 5G in connections will come in, analyzing what's going on in the car, or analyzing for threat detection, anomaly detection, things like that.

Also see

20190124poliakbill.jpg