EP39: The Software Defined Car: The Death of Freedom

Posted on By Jon Summers No Comments on EP39: The Software Defined Car: The Death of Freedom

John discusses his experiences at two recent presentations on the future of automobility. The first presentation by McKinsey focused on the concept of a ‘software-defined vehicle’ (SDV) – likening modern vehicles to connected devices similar to smartphones, and outlining six key areas for achieving success in this domain. These areas included hardware simplification, increased collaboration among manufacturers, leveraging cloud infrastructure, and shifting the focus from traditional engineering to software-driven engineering. The second presentation by a company called Aicas emphasized AI’s role in achieving interoperability and managing data within the intelligent vehicle edge, highlighting the importance of selectively recording useful data to avoid latency issues. Both presentations indicate an automotive industry pivoting towards making money from data rather than traditional car components and underscore the industry-wide challenge of securing and managing this data.

Down the road waits misery … Why cannot death just set me free?

Notes

Jon Summers is the Motoring Historian. He was a company car thrashing technology sales rep that turned into a fairly inept sports bike rider. On his show he gets together with various co-hosts to talk about new and old cars, driving, motorbikes, motor racing, motoring travel.

Transcript

[00:00:00] John Summers is the motoring historian. He was a company car thrashing technology sales rep that turned into a fairly inept sports bike rider hailing from California. He collects cars and bikes built with plenty of cheap and fast and not much reliable. On his show, he gets together with various co-hosts to talk about new and old cars driving motorbikes, motor racing, and motoring travel.

Good day. Good morning, good afternoon. It is John Summers, the motoring historian. This episode ostensibly is gonna tie together two different presentations that I went to coincidentally, you know, back to back. I. And I thought, you know, just because they fell, you know, one on a Wednesday night, the other on a [00:01:00] Thursday.

It was, it was literally, it was just a couple of weeks ago and, and I thought it seemed worthwhile. So, you know, it seemed natural to compare them together since, especially since both of them look at the future of Autumn Mobility. I mean, guys, it’s over, right? Let’s make no mistake. It is over. The first presentation was by McKinsey and it was about the software to find vehicle.

Fundamentally, this is no longer the car as a thing of independence. Now it’s like your phone, it’s the edge of the cloud. It’s a connected device.[00:02:00]

It is only now when I say it that I realize how much I absolutely just loath the thought of that.

The second presentation was, uh. One of the ones at the Nordic House organized by the Society of Automotive Engineers and, uh, SVE Beaker. Thank you again, sve for, for inviting me to these events. And this one was, uh, you know, again, one of these companies who are in the valley because there is a massive automotive presence in the valley and [00:03:00] they’re in the valley trying to understand the zeitgeist, trying to meet people, trying to understand and be part of the buzz, which is auto mobility in Silicon Valley.

So to be clear then these two presentations are thinking about the future of auto mobility in in two different ways. McKinsey are thinking about the future of auto mobility in a sort of strategic kind of way, and in a way, which is where they’re trying to look further and with greater perception [00:04:00] than other guys.

So, you know, they’ve rocked up. We’ve got the telescope, let’s look through the telescope. What can we see through the telescope? It was applicable that that one came first. Really? And, and you know that it’s McKinsey who are this like outside consultancy firm who, uh, just getting fucking everywhere recently.

If, and, and, you know, booty G the former transport secretary there, he’s a, a, a McKinsey guy. You know, my, my wife works for one of the magnificent seven corporates and, you know, her life is full of McKinsey people and McKinsey Method and you know, at McKinsey we do it this way, kind kind of thing. You know, these guys.

My students at Stanford, you know, they aspire to be a, a, these, uh, uh, uh, uh, the top consultancy firms and, and, uh, you know, in my day it was like Deloitte and Touch and Arthur Anderson. Now it’s, it’s Bain and, and especially McKinsey. So if [00:05:00] anyone’s able to tell you what the future’s gonna look like, it should be these guys, right?

Because these guys, uh, making millions outta the fact that they’re being hired. By the car companies to tell them what the future should look like. So these guys are literally, literally, you know, these guys are Isaac Asimov or Arthur C. Clark. They’re actually writing the future. ’cause think about that, right?

I used to feel like stuff was invented, but nothing’s invented. You know, it’s not like, you know, a mad scientist coming up with some magnificent invention like, I’ve done it now we can travel through time. No, it’s not like that at all. It it’s teams of software engineers and if the widget needs to watch it, if you spend enough money, the software engineers are able to develop the product that can do that.

So literally when you are designing a mainstream product like an iPhone or a new Tesla [00:06:00] or a new Ford Transit, you were literally. Designing the future because these are gonna be produced in their millions and sit around on every street corner or in everybody’s pocket, or, you know, the bios the unique way of starting the Model T Ford or the, you know, the Apple way of doing the phone versus the Android where the phone, this is something which is gonna be used by millions and millions of people and seep deep into the cultural zeitgeist.

Look, and that’s really what we’re talking about here because I’ve been watching a lot of Dave Freeberg road trips, a guy from Hot Rod Magazine, and, and he does these sort of road trips around Southern California on, uh, route 66 as part of, you know, his attempt to launch his own YouTube channel. And, and, uh, I find these kind of road trips really enjoyable to, to sort of ride [00:07:00] on along with him.

And I don’t know why I was talking about him.

What Free Burger does is celebrate. Automotive culture and what Freeberg does by visiting parts of Route 66 with their quintessential gas stations, with the portico on, on the front of them. This is, this is the modern Bailey castle of the American West, right? This is is a piece of architecture which speaks so much to the formation of the nation.[00:08:00]

And is just overlooked and forgotten and not really appreciated. Not in the case of the Martin Bailey Castles. There’s no English heritage, no equivalent of English heritage looking after these kind of properties. And that’s why I hesitate to call it work. This sort of folk archeology, that freeberg indulges in where you go and you look at the things and it’s very nostalgia driven and it’s kind of a historical, sorry, Dave.

I love the work, but you’re not, you’re clearly not a trade historian, are you? You’re a, you’re a Hollywood storyteller rather than somebody who’s thought properly about history. ’cause Dave Free Berger doesn’t like the graffiti on buildings. And whilst I understand that he doesn’t like the graffiti, I was looking at a, a, a piece of architecture the other day that’s had graffiti on it.

And I, I feel like you have to be able to look. Beyond the Graff, you can see the graffiti for the art that it is. If it’s good, sometimes it is, usually it isn’t. There’s something in [00:09:00] that. And the idea of graffiti is an evolving art form that’s really cool and that young people want to do it. That’s really cool as well.

So, you know, there is something to be said for graffiti is, is what I’m saying. And, and while I prefer, you know, free burgers, no graffiti on these buildings just to see the. Decline to see how it used to be to what he likes to do is that thing that that headmaster of mine in elementary school used to like to do where he used to imagine you were in the Roman bath when you were standing in the field, in the reign of, of Middle England on the site of, of the ancient Roman villa.

And he would imagine that you were in the hot bath and you could, you, you could, you could feel it. Well, Freeberg likes to do the same thing. He likes to imagine the cars that stopped here, the hot rodders who stopped on the way out to Al Mirage at this long forgotten rest. Hold it. It’s is really a, a, a, a interesting, uh, interesting how similar that that kind of [00:10:00] of history and freeberg kind of folk history is.

Although, as I say, Freeberg struggles with the notion of, in, in one episode I watched recently, he talks about how is it, I think that graffiti’s cool if it’s old, but not if it’s new. Failing to recognize that the paradox is you, Dave. The paradox is you, yourself, your attitude to it. You can, you just happen to like some things and not others.

That’s what’s uh, uh, at work here. It’s interesting how being historian gives you greater insight just ’cause you’ve thought about things in a different way. So look, Dave celebrates the past and I try and celebrate the past, but recently I’ve talked a lot about the future and I’ve talked in recent pods an awful lot about new cars, even though they do the job well.

But I don’t really give a shit about them. This is quite an important presentation though, and quite an important topic. [00:11:00] Um, because of this combining of the McKinsey vision with this very, very narrow vision of this one single software house presenting their suite of, of products to an audience of Silicon Valley people whom they are looking to get into bed with.

Let’s not beat, ran the, the bush. You know, that’s why the CEO is there doing the, doing the presentation. These guys were called acas Cas. Any presentation you go to at the moment will talk about the volume of data, which is being generated by modern cars, you know, contemporary cars, let alone the, the kind of systems that, that are coming in future and, and, um, nobody’s looking after this data properly.

Fundamentally, nobody knows what to do with it. These guys are offering half a dozen [00:12:00] opportunities and. Fundamental, you know, some kind of, uh, uh, tools in some cases for products, but in some cases just some tools around how you can take advantage of all of this data. So in other words, McKinsey gave you the big picture and then this software house showed you the sort of sinus of what the financial and technological infrastructure of these new software defined vehicles, data defined vehicles.

But software defined vehicles was the language that McKinsey used. So that’s how we’ll we’ll call it, they called it the SDV. It took me a while to, to figure out that, that acronym, it’s funny when you figured it out, it’s obvious, but it took me a while to to, to get to it. So this is the software defined car.

[00:13:00] V and well, you know, the software defined car versus data management and packaging. In the second presentation, you know, to read from the notes that I made before I started the presentation, the bottom line here is that cars part of connected system. In future they are, as I, so I think at the top of the show, they are edge of cloud devices like your phone.

So there’s no independence and there’s no freedom. The freedom of the open road is gone. I mean, I don’t know what you need to do, like get out and a Chevelle now. Like why you can is really, that’s really the position that we’re in. I mean this, uh, fellow that presented at McKinsey, he was a charismatic, likable guy and like the auto live.

Girl, Hannah, he [00:14:00] claimed to be a car guy, but you know, he told a story about fitting speakers to an E 36 when he was a teenager. It was on the tip of my tongue to ask him if I still had the car, but I knew he hadn’t. And this bloke came and talked to us for an hour or so about cars. He made a point of getting to know us.

As I say, he was charismatic. I liked him. It was a good presentation. He talked to us about cars. When I left, I’m always the last person, one of the last people to leave. I was chatting with another person about formula, one of all things, which is really interesting conversation, interesting bloke that I, uh, I I met.

But anyway, the presenter was still waiting for his Uber. He talked to us about cars and personal mobility, but he himself is waiting for his Uber. And the poor bastard’s been standing in the rain in his nice suit for 10 minutes waiting for the Uber to come because guess what? These new [00:15:00] fucking solutions don’t work.

They don’t work.

I, I’m not sure I’m, I’m increasingly feeling the, there was a. I, I felt that the pinnacle was the turn of the century, but I actually feel that there was an extended pinnacle into this century, which ended when, about 10 years ago, when manufacturers were forced to reintroduce emission standards. And when we first start, you know, really tough emission standards.

So when we first started getting, um, wet belt engines and, and these small capacity turbo engines,[00:16:00]

the McKinsey presentation was called the Mobile Destiny of 2030, winning the race in the Software Defined Era. I put SDV in quotes up above. And they’re like, you know, black piece of paper is like something important. So we talked about, uh, what software defined meant and how that’s designed, how that’s development of the product, how that’s upgrades on the fly, what they call OTA, like over the air upgrades.

How in EVs, that’s powertrain management. When we talk about software defined, we’re saying that, that the software [00:17:00] infrastructure of the vehicle is the bread of the pizza, right? It’s the most important thing about how. The product is designed. So you would say, is it like the steel of the car? Well, I, I, I don’t think it’s like the steel of the car, but I think it’s because, you know, everyone makes the car outta steel.

But I think if we think about what has made the German product stand out up to now, that there was a sort of better engineering integrity about the way that, you know, you didn’t need to be a car guy, just sit in A BMW and know it drove better than a Chevrolet. Yeah, sure. There were some Chevrolets that drove great, but by and large BMWs drove a certain way and you paid more to drive A BMW then you did to drive a Chevrolet.

That sort of differentiation that used to come from the powertrain and the interior [00:18:00] and, and all, who knows, some of it may still come from that, but I. In a software defined vehicle, the vast majority of that differentiation is going to come from software defined elements, which are going to be delivered over the airwaves to a product which has that cap capability built into it already.

Now, you may remember the BMW heated seats debacle of a few years ago. This is where BMW was selling cars that had the heated seats in them already, but you needed to pay an upgrade online to be able to use them. Well, people were pissed off that they work at the moment. You just need to enable them. I remember having a conversation with Mark Gamy who comes on the PO with his brother about [00:19:00] it, who was absolutely incensed of the notion that, that you should have to, that you should have to pay for this some.

BMW withdrew them. It’s not gone anywhere, but, but this McKinsey fellow is basically, that is gonna have to happen. People are just gonna have to eat that shit sandwich because, um, this is how people are gonna make money out of cars in future. He talked about, I’ll put this in quotes, extending the life of the hardware by making the software updateable on the fly.

So he almost seemed to be, I found myself. Thinking about was announced quite recently that the B 52 is gonna have their life extended at least until 2050 or some kind of ridiculous date where, by which time the airframes are going to be older than the people crewing. And of course, yeah, the airframes are, but [00:20:00] the software infrastructure and the engines.

There, those things are gonna be completely state of the art. It’s just the airframe, the building, if you like, the architectures the same and you know, it, it, it’s interesting that planes can, can work like that. And it’s interesting the, uh, to think that cars are, are gonna be able to, to, to work like, like that as, as well.

And, you know, we are comfortable, you know, when you buy a car, you are comfortable with the idea that you might, you know, chip it for a bit more power, might you, and you are comfortable with the idea that you might fit aftermarket wheels, you might fit aftermarket suspension, you might do some things to, to personalize it.

So although James Gamy was highly offended at the idea of having to pay for heated seats that were already fitted to the car, the notion of paying extra for he heated seats is not. What upsets him, it’s [00:21:00] just having to pay for it when you’ve kind of already got it. There was just that kind of hub to get over.

But you see, when everyone’s used to a over the airwaves update on, on their car, they’re not gonna feel like it in the same kind of, of, you know, Luddite way that, that James did. I’m sure you won’t mind me saying that, Jamie. No. Something he talked about was how, there’s always a question about how much compute is actually needed on the edge of the cloud.

In other words, you know, what you want to do is just have like a green screen set up, but actually with cars you have to have some processing power there. On the edge because connectivity fades and there’s latency and it’s a high critical situation where you can’t have, you know, latency in, in decision making with, in some cases, if the car’s driving on the freeway and it needs to decide whether to swerve or not.

There can’t be any, can’t [00:22:00] be needing to connect to some computer over the internet. And if there’s, you know, if there’s the, you know, the spitting wheel of doom, you, you see what I mean? The car has to have some computing power actually on board, so it could make, it can make good decisions. And what the McKinsey fellow termed this as, as he termed this, as a centralized yet distributed system.

So we had one slide that had six areas which were gonna define whether or not. You were gonna win or lose the race to a software defined automobile. It’s not an automobile at that point, is it? It’s just not. I mean, if a Tesla, whereas I wrote an article many years ago where I compared old Shavel with a Tesla and just said that, you know, it’s just less of what I understand a car to be.

It’s just, [00:23:00] you know, less, I know it’s better as what we define a car to be, but it’s not what I, you know, understand a a, a car to be. The Chevelle still fulfills that definition more than than a Tesla. It’s Tesla’s an awesome piece of technology. It’s just not a car, as I’ve understood it. That was my attitude, sort of 10 plus years ago.

Now I see that they redefined what the car is. Right now. Everyone’s using them. They’ve just redefined what, what, what the, what, what the car is, where these, you know, software defined vehicles of the future are gonna do that more, more and more.

The notion that the automobile is a [00:24:00] place of personal independence and freedom, I think is gonna disappear. I think it’s just gonna be another environment where you are connected and it’s just gonna be another device like your phone. Yes. It’s sick making, isn’t it really is. You know, you, you want to try and.

Put a brave face on it, but

how you can talk enthusiastically about an E 36 BMW and then enthusiastically present this like computer on wheels. Shit, I just, I just,[00:25:00]

anyway, the first factor in determining whether or not you’re gonna win the race to NT cars and invent these horrible mobility, mobile phone, meet cars, things. Um, the first area is gonna be, uh, hardware platform simplification. And that’s a necessity, not a, a, a choice. Apparently Rivian have 17 ECUs in. I remember reading back in the days when I used to read car magazines, so least 10 years ago, that you know, the wiring in a BMW seven series, if you stretch it out and ran it all around the world or ran it, it would stretch around the world.

You know, there was that much bloody wiring in them. So you need to simplify that. And the parallel was the, like your phone, now your phone does everything. Used [00:26:00] to have mice and printers and cameras and all of that stuff. And now you don’t need calculators. You don’t need any of those devices all like tied together on your phone.

And that needs to happen in the design of automobiles and mobility devices. Or it will happen. The second area is that there needs to be more collaboration and and less competition. If you listen to my, uh, leadar presentation, the haw Wind and, and Leadar presentation I did a few months ago, one of the things that I was just gobsmacked by was that in this bleeding edge way, the autonomous vehicles are perceiving the world, which is are instead of companies pooling the research or sharing it or something like that, there’s three competing German makers and they all have [00:27:00] different protocol and they’re all competing with each other.

And each Volkswagen and BMW and Mercedes, they all are desperate to preserve some kind of sense of identity and competitive advantage. You know, we’re doing the work three times instead of just once it’s, and that just seems really bloody absurd. It seems to be how we did things in the 20th century, not how we.

Need to do things now, it’s so much wasted labor. Even if you think we need autonomous cars for three people to be working on the same piece of technology, which Tesla and Musk feel is not even needed at all. It just seems very, very, it seems likely to be misguided and wasteful, put it that way.

The, the [00:28:00] other example that, that I wrote down here around this more collaboration than, than competition area is that GPS is used for maps and, and, and safety around the world, and, and that’s the sort of model that we need to work towards. There needs to be that kind of shared infrastructure. There

again, right? I mean, BMW might be able to navigate it, but Chrysler Reno, like, are they gonna be able to navigate this kind of world against the Chinese? I mean, in 1960 it was hard to perceive that the British motor, indu motorcycle industry was going to be [00:29:00] gone in 15 years, gone. The Australian motor industry ended in one year period.

Within the last five years, there was all three. Uh, the big three were making cars there. One minute and now nothing at all. The sun is really setting. I mean, I don’t wanna keep harping on the, it’s over me, but guys, it really bloody is so, yeah, so, so one simplification of the hardware platform. So instead of what American cars were like, the fifties, a simple platform that’s defined by, you know, the features that you have in it.

Everyone has the same iPhone, but you know, some of you have different apps in it and you pay for more of the apps or for security or, or whatever. That’s, that’s how these things are gonna be differentiated for these things to work. We need not to be doing what the Germans are doing with Leadar. [00:30:00] Everyone work there needs to be sharing of the data.

Or at least of standards, you know, protocols, the way that we’re gonna, you know, Android and Mac have what was fundamentally a Windows infrastructure, don’t they? There’s that kind of basic understanding of whether we’re gonna drive on the left hand side or the right hand side of the road, or whether we’re gonna steer with wheels or tillers.

That’s the kind of thing that, that needs to be, uh, figured out here.

I can’t read why I wrote for number three, but didn’t write any notes around it Anyway, number four is cloud will be the engine powering automotive. [00:31:00] He used this phrase, hyperscalers, which I’ve not used before, but was used in the second presentation as, as, as well. And by that they mean, you know, Amazon web services or basically people who give you compute power on demand.

So this hyper scaling, getting that right, having the right partner with that, since it’s not about the edge, it’s more about the center. Making sure that what you’ve got at the center works properly is going to be key. Right? The cloud infrastructure is what’s important. Far more important than what the design of the devices on the edge of, of the infrastructure.

In, in, in other words, it is important to be Verizon or the network provider as it is to be Samsung or Noia. If they’re still, you know, the, the, the blokes that actually make the devices. That you can see[00:32:00]

engineering must shift. I wrote, this is really interesting because this is a change that’s reflected in my wife’s place. My wife works at a, you know, magnificent seven company and it has been a shift where it used to be the product called the shots and sales sold what Product built. And you know, I’ve worked at Valley companies that were like that as well, where the engineers were all the C-level people.

So, you know, they built what they wanted to build and they hired people to sell it. And then they were like, but why aren’t you guys able to sell this stuff? And it’s like, because you, you’re fucking building what you want to build rather than what customers actually want to buy. Right, and that’s what’s happened is that sales have come along and said, no [00:33:00] post pandemic sales have come along and said, no, you need to build what we wanna sell.

Don’t build this shit that we don’t want, that our customers don’t want. Sales is king, not product. The same is apparently called the McKinsey needs to happen in automotive. That, that traditionally it was the motor and chassis people who called the shots and decided how and what the car was like. But now software’s king.

Now the motor and chassis, they don’t really matter. That’s the real indictment, isn’t it? The motor and chassis people don’t matter. And if you think about it, you know, all the time it was my wife’s place that sales were taking over and, you know, senior people were being fired and their jobs were sort of just being dumbed down and given to, to lower level people because.

You know, it’s just changing. AI’s being used to do some things. The organization is just, you [00:34:00] know, instead of doing everything to the nth degree, you’re happy to do it to the less than nth degree, but for a better price with a, you know, lower caliber person. That’s where we’re going with this, around the notion of people are used to the idea of buying a chip to like improve their performance or, you know, the car’s performance or some alloy wheels, but they don’t wanna pay for the heated seeds that they perceived as already being built into the car.

He said that he felt like makers needed to focus on what the delightful, what were the delightful experiences that you were gonna deliver to your customer. So the final one was, uh, the, in order to, you know, make the shift towards software and away from. Traditional engineering. And if you think of it, by the way, ages ago, I think in the, I went to a presentation, the zf, uh, the [00:35:00] transmission, the German transmission people did.

And, and I realized that there were two Zfs. There was the old ZF that made the gears, and then there was this new ZF that was, uh, called zf, but it was actually the software house that they’d acquired because EVs don’t need normal gear boxes, do they? They need like programming and software. So ZF had completely changed what they needed to do, and it basically turned into a software house.

And they’re, you know, the guy that I was talking to who was their head honcho in America, was the CEO of the whole software house that they, you know, the software, boutique software house. They bought a few, you know, a year or so before in order to make this pivot from making cogs to making software. So, you know.

You really need engineering talent for this. And I underline that because, you know, that’s, this is what McKinsey was saying is that, that, you know, you’re just not gonna make this [00:36:00] pivot unless you get the, the right engineers on board. And, and a lot of them, so the revenue model here is business to business data sales.

I, we talked a little bit about car advertisements and how already they’d pivoted away from talking about, you know, power and speed and far more They were about the journey that the consumer was on. That was the quote that he put in it. And I, I found myself thinking of these sort of Subaru adverts do this very well.

If you were, you, you feel. The overwhelming feeling is the sort of person who owns a Subaru and you wanting to be that sort of person that owns the Subaru. The Subaru is just like the stage upon which the little like, you know, family, dog loving, you know, [00:37:00] DEI family is playing out against. The example he gave was that the car will be like a TV where you buy streaming programs.

I can’t think of anything worse

if you’re gonna sell the car and then do over the airwaves upgrades to it, you know, in the way that you do a laptop or your phone, the device, the car has to have. Processing power headroom built into it.

The software also [00:38:00] has to have this sort of headroom built into it without one to jump ahead. One of the issues that the second presentation highlighted was the way that if there’s a problem with the development of a particular piece of software with a car maker at the moment, they’ll go to a consultancy and that consultancy would deliver a point solution and then.

The car company will move on and design another model and maybe, you know, hire the same consultancy to do a different or another point solution. And these point solutions, they may be written in a protocol that means that they’re just not updateable, they’re just cast in stone. You know, they were a point solution that worked for that particular application, but that’s all they were.

There was no kind of, you know, future proofing, uh, around them. It’s not just they need extra processing power. It’s also the kind of this way in which the software’s gonna have to be [00:39:00] engineered to cope with the fact that these, that, you know, the software’s gonna be coming out every couple of months, but the car’s gonna be on the road for five, 10 years.

So we, we had a sort of slide about what was a software defined vehicle. It’s a business model. You know, you’re making money, selling the data, not. Your car’s doing better gas mileage than the other guys, or has got more horsepower or, you know, as nice as styling.

I also wrote something about customers, the way the customers are, but I can’t remember what I wrote, like I read my own writing. And then the third sort of characteristic of the software defined vehicle was the actual hardware itself.[00:40:00]

Next, he talked about security, how you have to create a sort of DMZ around systems like breaking. He had this fucking amusing metaphor. Where we’d add like a pasta dish at the restaurant we were in, where we were doing the presentation. It was, it was over in Oakland and it was looking out over the harbor and it was, uh, you know, it was at nice kind of picturesque there.

And, uh, you gotta imagine this McKinsey bloke, uh, in his nice suit describing how each different heated piece of Italian pasta or ravioli or whatever it was, how each one of those was similar to the sort of DMZs that were required on cars. So in other words, if this one on the left here is your ice, you know, your in-car entertainment, [00:41:00] streaming music, all of that kind of stuff, that is something which you would allow over the air updates to take place, right?

Because ultimately if it doesn’t work, it’s annoying, but it’s not the end of the world. Whereas the breaks, that is something which. If you were gonna do over the airwaves updates to it, you would need a system, uh, ensuring that those updates took effectively and that the system was gonna work properly all of the time, for obvious reasons.

Just stop and think about that, right. The, the breaks. I mean, just in the, um, I did a poll a couple of weeks ago where I used the, the quote from 2001, A Space Odyssey. You know, I’m sorry, Dave, I can’t let you do that. Where the computer thinks it’s gonna, you know, the humans are gonna turn it off and it’s like the mission’s too important.

Dave, I can’t let you turn me off. The music, the [00:42:00] machine’s taking over, it’s the classic, you know, plot spoiler for 2001. A Space Odyssey there guys, by the way, sorry about that.

If every element of the car can be updated over the airwaves, when you get in it and you start it up and you back it out the driveway in the morning and it’s had a new software upgrade, it’s not even the same car as it was the night before. Is it, was it Star Trek or was like with the quote, it’s life Jim, but does we know it[00:43:00]

Well? This is quite alarming. This is why I enjoy going to these kind of presentations as well, because it, it, it makes you, you realize where we are because of course there are certain areas that they don’t know what’s happening. I, I wrote down, it’s not clear who owns the data and the data. We mean the data that’s being generated from all the cameras that are on cars.

Think about how many more there will be on, apparently the Waymo has 29. Pairs of eyes on the road. That’s what the auto live engineer was, was saying, 29 pairs of eyes. So think about all the data which is streaming off that, that car, and think about how powerful that is if you can, you know, big data it with all of the, um, all of the data for all the vehicles all around.

So it’s not clear who owns the data. It’s not clear who should store it, who should pay [00:44:00] for it to be stored. It’s not clear how it should be stored, you know, whether it should be nearline or that kind of method of storage, which is, you know, slower to get to. Gosh, this is reaching back into my own software sales past of, of where the things are online or near line, as we used to say, take backup.

Tape backup. That’s right. Youngins. Some motherfucker used to come to your premises and pick up tapes and take them off site, and that was your disaster recovery. If your factory burned down, you bought new computers, and then the dudes showed up with the tapes and your new systems always loaded absolutely perfectly without issue.

The data issue [00:45:00] is huge. Said the McKinsey guy. Then he left and had to wait 15 minutes for his Uber.

So the second presentation was by these guys acas, ICAS. Nothing to do with ai. It’s just a coincidence. The CTO and founder is named James Hunt. Faintly ironic about James Hunt being involved in the dein inventing of not just driving, but the whole, what happens in Vegas stays in Vegas. Freedom of the open road, freedom to.

Grab a girl’s ass without there being any repercussions. James Hunt, the motor racing driver personified that, didn’t he [00:46:00] personify that Playboy era of the 1970s, the freedom to do that kind of thing. Right or wrong, you know? ’cause my freedom is a, you know, another person’s oppression and all, you know, I, I get it.

I’m not advocating for that. I’m just saying in the 1970s, we’re a different place. And James Hunt was the personification of a way of living, which is far too misogynistic to be acceptable at the moment. And it’s ironic that, uh, this blo who’s invented this company with this technology That’s, yeah, the opposite.

I, I made that point, didn’t I? Do I need to keep hammering it? I’m not sure I.

So the acas presentation was called, is AI Ready for the Intelligent Vehicle Edge? I think about that. [00:47:00] There’s a lot going on there, isn’t there? And the guy talked a little bit about the title in the presentation, and it’s only now reading it back that I, I realize that, I’ll say it again to help you ruminate it.

Is AI ready for the intelligent vehicle edge? The fellow presenting was called, uh, Johans, Beerman German, president of the company, company based in Borough, which is a, a town, uh, not far from, from Frankfurt, where we actually have some friends who live. So that was, uh. It was a, a weird coincidence, but maybe not, you know, because really, um, this is the sort of automobile of Silicon Valley, you know, Ben’s and Kta and the first car and all of that was not far from, from Carl’s Ru.

So were so interesting, uh, that, that there was that connection. So he began his presentation with a story, and I thought it was illuminating for as usual different reasons, for the reasons that he thought it was interesting. But I love an anecdote. He [00:48:00] began with it. So, uh, you know, let me, uh, let me do it justice.

And it was, it was that he’d driven a cyber truck. Um, he had friends up in Tahoe or somewhere like that and he’d, he’d driven a cyber truck in the time that he’d been here in, uh, in California. He, he, he, uh, was surprised that the size was manageable, which of course is a tribute to how much space there is in America versus in Europe.

He’d described his first experience using full self-driving says he felt safe. Which is interesting because that’s the experience I have watching people ride in Waymo’s for the first time and then be like, whoa, it’s actually a really good driver. But, you know, I know that feeling of bemusement is still that I had the first time I rode in it.

That’s only about her, you know, nine months or 12 months old or something like that. So I do remember how it felt to be like, wow, it can drive. This felt, so, this full self-driving really works. The other thing he [00:49:00] said, and I thought this was really interesting because this speaks to the kind of thing that my students say when you talk to them about their interest in car design and product design and marketing and why they’re doing a class, which is called Tales to Design Cars Buying a, a surprising number of, of people will, will, will talk about wanting to create a better situation for, you know, certainly a justification for autonomy is that people who are not able to drive are still able to lead independent, meaningful lives.

And Johanna’s here, I’m not, I think that’s how you pronounce it. Apologies sir, if, uh, if I got it wrong. ’cause your English was excellent. Obviously my German is, uh, really not, not there. Um, but look, the, the, the bottom line is that he said, I thought of my dad. I thought of him towards the end of his life. I thought, how much richer those later years, [00:50:00] those last years of his life could have been, had this technology existed.

I met somebody at Stanford some years ago, really impressive lady. She was interested in car design. You know, she was one of the first people that I’d had a candid conversation with at Stanford about the fact that, you know, they didn’t care about Formula One, but they were really interested in car design from a completely different perspective, from the, the angle the the eye had had come from.

She had two sets of grandparents, each of whom lived in, they lived in different parts of the country, but they depended on, uh, the automobile for a sense of community, for their friends to, you know, different lifestyles. But, you know, the, the point was that they were. Desperate not to lose their driving licenses.

And if they lost their driving licenses, they felt like that was gonna be the end of their lives. ’cause they couldn’t get out and see people and socialize and, you know, because America’s the distances are huge and, you know, all, [00:51:00] all, all of that stuff that we, that we already know. So in other words, probably, probably 10 years ago now, she had that conversation with, with me and I realized that autonomy was a lot bigger than me.

I like to drive, you know, I like the feeling of a car in the Benz. Um, it was a lot bigger than that. It was really about it improving people’s quality of life. And I know I was cynical about that in the auto live presentation. ’cause the, the safety thing, oh that is just so, uh, uh, you know, sends chills down my spine.

But this is different from that because this is. The technology, the autonomy, actually delivering independence. It’s extending, it’s making people’s lives more meaningful. It’s saying you don’t have to sit at home saying that you can, or you don’t have to do like virtual chats with your friends. The bloody computer car [00:52:00] can take you out to church or the bar or bowling or pickleball or you know, whatever the devil your poison is.

So I thought it was. Interesting, his perspective on technologies, which, although he has this, he, you know, works for a leading German technology, software technology house. He felt, you know, he was like, wow, the fell full self-driving actually works. Like, wow, the truck’s not too big. You know, he had a kind of a, a, a, a very European outsider kind of reaction to, to the cyber truck and to, to full self-driving.

So that was really interesting. But what that made me realize was that it’s not about being forward or behind, is it? ’cause I remember going to England some years ago and everyone else, everyone in England tapping their card and me needing to slide it or something like that, and not having the tap [00:53:00] functionality.

You know, figured out. So I don’t wanna somehow imply that, you know, oh, we in California are so forward with, uh, with our, you know, dangerous, sharp-edged, ridiculous, huge cyber trucks. I’m not trying to suggest that at all. I’m just saying that the comfort level with full self-driving, I found that really interesting that the German, even a German working in the industry didn’t feel like that about it.

What he said was, for a German, this is very advanced. I I just put the, we just don’t test properly. You know, we just bloody beat at testing. ’cause we, ’cause we are, if you look at the Tesla accidents, I don’t wanna fall down this rat hole, but if you look at the accidents Tesla’s had, it’s where the full self-driving technology has not perceived the world correctly.

But John, but John, are you seriously telling me that full self-driving as it stands at the moment is. More dangerous than actual drivers. More drivers would’ve been killed than full [00:54:00] self-driving. As it stands, and that is true, isn’t it? If you think about it, that the computers aren’t perfect, but they’re still better.

Than most of the people than you know, all of the people all the time. They’re not perfect, but they’re better than, than we humans are. And, and I think, you know, we need to, to recognize that. I feel less like Elon Musk is beta testing full self-driving software on us. I think it’s less beta testing. I just think it’s like software.

And we know that all kinds of software is, is buggy and that the method of doing patches and patches and patches and patches, this is, it’s just you don’t need to be a computer scientist to see how, you know, a bridge that you are constantly having to patch or a ship that you’re constantly having to patch holes in the side of.

That’s it. Just, you know what I mean? The, the, the, I mean, yeah. Anyway, that is a whole separate, uh, separate rat hole, isn’t it? [00:55:00] Let me actually talk meaningfully about acas. The company’s 20 years old. And they’re about products, not services, and they operate like a licensing kind of model. I don’t know if, does that mean they license their own technology or they license other people?

Sorry if that’s not clear. It wasn’t clear to me automotive and QA in other sort of similar areas. So one of the things that he highlighted was, you know, luggage belts in, in airports, they have lots of customers. I mean, I took a photo of the slide, but you know, they’re a significant player in, in the space.

So at the moment, the kind of thing they do is over the air updates. OTA as the McKinsey guy taught me the, uh, the acronym they do data and software lifecycle management. So what that means is making sure that you know that this stuff’s backwards, compatible, that, you know, um, you’re not, that, yeah, that the stuff [00:56:00] is backwards and forwards, uh, compatible.

They do cloud and edge. Product. This is dense, but it bears applying your mind to it because it shows how this guy thinks about his business. Um, it’s quite, quite dramatic really. Interoperability is an Internet of things challenge. Is ai, the silver bullet, in other words, making all of these different protocols and different software products all talked together properly is like a mind-boggling internet of things challenge, but AI with its capability to do many small tasks very, very quickly and hopefully pretty faultlessly, um, certainly better than human beings.

Is this the silver bullet that’s gonna enable us to, to, you know, deliver the kind of autonomy that everyone wants? You know, that that was what we were sort of, of floating around the [00:57:00] idea of, and by the kind of autonomy that everybody wants. I mean, you go to a rock show, you get plastered at the rock show.

You leave the rock show, you fall into your car and you say, take me home to San Francisco. And overnight while you sleep off the beer, the car drives you from Los Angeles to San Francisco, and you step out of the car in San Francisco and grab a cup of coffee, walk into the office. Jobs are good. That is the vision.

Until autonomy can be that disconnected, you know, you’re gonna sleep overnight, undisturbed comfortably, and you can be drunk in the car. It is not going to require you to take over until that level. I dunno if they call that level. I dunno whether that’s level five or autonomy or quite what it is. That’s, that’s my measure of when true autonomy comes is, [00:58:00] is, is when you can do that.

He identified five ways AI can help with interoperability. Citing as examples, stellantis Ford and Volkswagen. He said that they have multiple platform software platforms that they’re catering to and just to ingest the data, you need interoperability, you know, just to. Take it on board without you do any kind of analysis and, and scrape or, you know, try and package it up and sell it.

This is that thing that I talked about earlier in the presentation where, because you need the interoperability, uh, car makers historically have asked system integrators to do piecemeal solutions, but then those piecemeal solutions aren’t future proofed. It just becomes really, really messy, and you’re in that world of patches.

On patches. We [00:59:00] can standardize this, we can platformized this. He said at the vehicle level, they can develop methods of saving only the useful data, not ev or not all the data that the vehicle produces. Think about that, right? Because what we’re saying now is that. Waymo Cab, according to the Auto Live engineer, 29 pairs of eyes on the road.

And at the moment, the Nvidia onboard and, you know, the Waymo data crunching back at, you know, Google HQ there. That is what’s allowing the Waymo to make good decisions. What this guy’s software is gonna do is empower the AI to not record certain things or to place it in a, in those, you know, never regions of, of computers that [01:00:00] only engineers understand where, you know, you can extract them.

Do you see what I mean? But you are, you we’re fundamentally given the ai, the chance to censor itself, censor its own behavior. Yeah, we need to strike a balance between edge and cloud. This is a more sophisticated way of saying what the McKinsey dude was, was talking about. Whilst the cloud infrastructure’s the most important thing, the product on the edge needs to have some compute, and this, the balance between edge and cloud.

Is, is that expressed in, in another way? And, and hence again, this term of, of cloud hyperscalers. And, and it’s interesting that whilst they used different terms, you know, the software defined vehicle data, you know, packaging, they, whilst they were approaching the same island from different angles. Some of the things that they said were the same and this cloud hyperscalers was, was [01:01:00] something that I thought, you know, in other words, we can’t do the data required for AI without the cloud hyperscalers.

Right? Which is presumably why the street sees so much value in the Magnificent Seven because, uh, of the level of compute that’s gonna be required to deliver the business solutions of the future. Johanna’s highlighted the pandemic as being the time where cars really became more software defined. And the way that he framed it, and this is thinking like a businessman, I talk to VPs now, not directors at the OEMs like I used to.

He says that the sales cycle is extremely long and that they need to be embedded early on in the process. You have to have the rigor to make sure that you do the right thing. And I wrote down afterwards that I felt like, you know, kind of Tesla haven’t, but nonetheless, you [01:02:00] know, I do feel like the full self-driving it is better than most of the people, most of the time.

Is it faultless? No. Where do we draw the line? You know, I think we have to draw the line when the cruise drags the woman down the street underneath itself. I think at that point you have to say at that point, Hal needs to bloody well open the portal doors. I’m sorry Dave. I can’t do that. I’m sorry. I’m homeless.

Drunk woman stuck in the wheel arches. I’m gonna drive along over you anyway. You know, clearly that can’t happen. But you know, the fact is a human driver. Yeah. Human drivers make more mistakes. Have I made that point over and over? I probably have, haven’t I? And if I haven’t made it over and over, I’ve now muffed the clarity of the point right there and then.

This is the trouble with my delivery style. What I should do is say it more clearly and then edit out all the bits where I said it badly, but I’m just not like that. [01:03:00] He made an interesting point that although V FAR can have lots and lots of technology in house, they still bought Rivian or they’ve invested heavily in Rivian for the data share.

Right. That’s what he’s implying that there is. It’s a land grab. It’s what’s going on for engineering. Talent. And, and it’s also that thing that, you know, looking back to the 1920s, we can see very clearly the, the automobile was standardizing and many small companies with different solutions were consolidating.

And that would happen in the postwar period as well, into a small number of, of very large companies. And this is to say that, you know, whilst they used to be like Oakland and Chevrolet, um, Oakland then became part of Chevrolet and, you know, adopted the Pontiac Moer. And then what Pontiacs were was slightly more at market from [01:04:00] Chevrolet, and certainly not a Buick or a Cadillac.

You know, there was, there was that differentiation taken place to achieve the AI and data interoperability. They create an abstraction layer. To automate many smaller integrations. That thing that we talked about a moment ago, the the way it does selective relevant data, the point he made is that however much processing and bandwidth you’ve got, it’s still critical that you understand what data’s important and what data isn’t.

To avoid latency, there’s a standards organization called coa. It’s interesting, right? ’cause that’s what SAE are. So it’s interesting, what engineers need as they’re developing their products is to create these kind of bodies that allow them to agree that, you know, [01:05:00] we’re all gonna use bolts that have the same kind of, of thread.

We’re all gonna use the same programming language. We’re all gonna use the same fundamental architecture. In this case, what he’s talking about doing is standardizing can data. So, so you know, the data that, that the car itself, you know, the language the car communicates with itself in if, if you like the local area network of the car, you know the hard.

To share data costs money, different departments, so you know, the suspension guys don’t necessarily wanna share their data with the navigation guys, even though it might be beneficial for both parties because they feel like they’re giving up some of their internet. Intellectual property. So there’s sometimes silos within car companies and, and even if, you know, they were enthusiastic to share the data, who paid for the [01:06:00] infrastructure for the data sharing to take place?

That whole stuff that I used to sell years ago, it’s the storage, but it’s also the safe and secure transfer and the safe and secure retrieval of the data. It’s more than, you know, storage is more than just storage.

I wrote in the margin that the guy looked like Harry Hill and had similar mannerisms. And I mentioned it because it became more off-put as the, uh, the evening went on. As I, at first I thought I sort of recognized him, and then I realized that this was Harry Hill’s, serious German engineering cousin. And the more I thought of him being like Harry Hill, the more the mannerisms became, you know, it’s literally, it could have been, he could have been Harry Hill’s, sort of straight brother, really straight man, brother.

[01:07:00] And one of the most interesting parts of, of the evening was when, uh, he talked about partnering with a company called NXP. And in other words, this acas software sits on a NXP box. And, and actually if, if you Google around, if you Google NXP, um, acas appear as a sort of, you know, partner. On the MX NXP website, one of the guys NXP, also based in borough, one of the questions from the audience was, was about whether or not NXP could exist without a CAS or a CAS could exist without NXP.

And what the guy was probing around was whether you actually, whether the OEMs actually needed the acas software or if they could just get away with just [01:08:00] using the the NXP. And the answer to that was that you can just buy NXP. You don’t need to have the acas. So, so the acas is, I wouldn’t quite say a software, I, I wouldn’t quite say a sunroof, but you know, it, it’s more something like all wheel drive.

It’s, it’s something which you need to decide quite early in the process of developing a car or which, you know, whether you’re gonna have the belt and braces or, or, or not. The data’s controlled by the customer, not acas. I mean, I, I wrote at this point in my notes that I’d realized that I was comparing the, the two to together and, and I felt like both of these presentations were about making money off data.

Not about making money from motors, seats, bodies, nice interiors, any, all of that is just not on the table anymore, just as your [01:09:00] computer has, you know, some importance as a device. But it’s the software that the computer’s running, which is really the important functionality and the important thing that you are ready to either pay for or not pay for.

The, the device itself, there’s a little bit of fashion accoutrement, uh, uh, about it, but, but beyond that, it’s just a disposable thing that you just replace every now and again. It’s certainly not something that you do any kind of maintenance on.

Data is controlled by the customer, not acas as, and I’m emphasizing that because the way that spend framed that quite amusingly was he said, ah, you guys do the plumbing, but you’re not responsible for what gets flushed. And that is an interesting analogy [01:10:00] because that’s sort of what we’re talking about with these platforms like X and, you know, meta or, or, or whatever.

And the, the level of, of censorship or mediation or whatever you want to call it, that we feel is it the platform’s responsibility to ensure safety. That’s what we’re driving at here. And, and, uh, they’re not offering a solution to who owns the data. They’re not offering a solution to who stores the data.

All their offering is a way in which you can package it up and make money out of it as a OEM. They’re selling a way for OEMs to make money. You know, they’re selling shovels in the gold rush of automobile makers rushing to sell data. This is, these are the shovels.[01:11:00]

Now, his version of, you know, the ravioli box being separate from the salad box, being separate from the, you know, salmon pasta box. Um, his version of that was z uh, the A zoned architecture. Well, he actually said a zoned architecture was evolving. So in other words, you know, that was, was already taking place.

There was already, you know, already the, the salad of breaking was totally separate from the in-car entertainment of the river early.

I was amused that both presenters used the [01:12:00] same graphic when talking about, you know, different systems on the car. And they got some slide that had a car in profile with dots on it. And, you know, their bullet points going off from, from the, from the dots. And I was amused. I was amused. They used the same image.

It’s a recurring theme in automotive presentations that there are a couple of images which seem to get used over and over again and it makes you realize how. Linear. The narrative has been, if we are looking at the same photos over and over, our thinking seems to be very, very linear. And these two guys had the same image, you know, coincidence.

Yes. But you know, maybe there’s something more to it.

So to that [01:13:00] point that you get up in the morning, you back the car out the driveway, um, and you’re set off to work, the car you’re driving is not the same car as it was the night before. It’s not right. It’s not because the AI is what’s updating over the airwaves. I. It’s not that, oh, you know, not like your laptop at the moment where, oh, it’s got like a new version of the software and you know, then you, you have to reboot and then you carry on.

It’s not like that. It’s the, it’s, so it’s actually going to make different decisions now than it did the night before. Better decisions now, I’m sure. I mean. Driving is about control, isn’t it? That is about completely yielding control and not even knowing you could get in the car and drive it, and the AI parameters could be changed completely.

It could have changed into Junior Johnson mode, and you’d been none wiser.[01:14:00]

So to sum up, then, acas had. These five ways in which they, they could help. I did take a photo. I might link to it if I can, but you know, if you’re listening to this presentation, it’s enough. I think you know, the takeaway is, is that structure to make money outta the data is being put in place, even as the securing and legal clarity around who actually owns the data and who has the right to do what with it.

Even as that remains completely opaque. The method to make money out of the data is taken shape. Thank you. Drive through

[01:15:00] black

down your soul to the God’s. Rock and roll

black. This episode has been brought to you by Grand Touring Motorsports as part of our Motoring Podcast network. For more episodes like this, tune in each week for more exciting and educational content from organizations like The Exotic Car Marketplace, the Motoring Historian, break Fix, and many others.

If you’d like to support Grand Touring Motorsport and the Motoring Podcast Network, sign up for one of our many sponsorship tiers at www.patreon.com/gt motorsports. Please note that the content, opinions and materials presented and expressed in this episode are those of its creator, and this episode has been published with their consent.

If you have any inquiries about this program, please contact the creators of this episode via email or social media as mentioned in the [01:16:00] episode.

Highlights

Skip ahead if you must… Here’s the highlights from this episode you might be most interested in and their corresponding time stamps.

  • 00:00 Comparing Two Presentations on Auto Mobility
  • 01:21 McKinsey’s Vision of the Future
  • 02:34 Nordic House Presentation and Silicon Valley’s Role
  • 03:34 The Future of Auto Mobility: Strategic Insights
  • 06:32 Reflections on Automotive Culture
  • 12:15 The Shift to Software-Defined Vehicles
  • 13:05 Challenges and Opportunities in Data Management
  • 16:53 The Evolution of Car Design and Engineering
  • 26:23 The Role of Collaboration in Automotive Innovation
  • 30:45 The Importance of Cloud Infrastructure
  • 40:05 Security and Over-the-Air Updates
  • 41:35 Reflecting on AI and Classic Sci-Fi
  • 43:15 The Data Dilemma in Autonomous Vehicles
  • 46:50 The Role of AI in Intelligent Vehicles
  • 48:03 Personal Anecdotes on Autonomous Driving
  • 55:03 Challenges and Opportunities in AI Interoperability
  • 01:01:18 The Future of Data in the Automotive Industry
  • 01:14:14 Conclusion and Final Thoughts

Enjoy more Motoring Historian Podcast Episodes!

The Motoring Historian is produced and sponsored by The Motoring Podcast Network

Podcast Tags:

Related Posts

EP6: Blue Thunder, Black Widows, Avant-Garde Freightliners and a Runaway Cavalier Podcast
SPECIAL: MILLE MIGLIA 1954-1957 & Beyond! Podcast
EP21: Hydrogen Trucks Through The Lens of a 9 Year Old Podcast
EP1: 1100HP Lucid Driven; Isle of Man TT, Our Man Still in Hospital Podcast
EP30: Automotive User Interface Conference, Stanford 2024 Academia
EP24: Stonehenge by Mokka; Saxon & Gixxers Podcast

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.