What Lies Ahead for Auto Industry in 2024

[Daniel Nenni]

Digital twin technology can shorten automotive development cycles.

What Lies Ahead for Auto Industry in 2024​

Trends in 2024 will demonstrate the need for extended collaboration at all stages in the automotive development process.

The race to develop the most advanced, secure and safe vehicles on the road has been on for quite some time, but we expect that race to gain speed in 2024. Global competition between automakers, consumer demand for advanced features, upcoming automotive industry regulations and new entrants to the automotive silicon market are fueling this fierce battle.

Despite this competition and intensified market pressure, there is still a need for extended collaboration at all stages in the automotive development process to manage the increasing complexity of vehicle systems-on-chip (SoCs), the evolving cybersecurity threat landscape and shortening development cycles.
Here are our top four predictions that we see guiding the automotive market in 2024:

1. Market Pressure Yields Shorter Automotive Development Cycles

Automakers are rushing to keep up with Chinese manufacturers who are setting the pace of development at levels previously unseen in the traditional market. Due to robust automotive safety standards and resulting testing needs, the technology in a brand-new car is likely to be five to seven years old. Chinese manufacturers are closing that gap to be closer to three to five years. Both silicon and software testing validation need to happen quickly to accelerate development, often in parallel with one another.

One way the industry is lessening the time to market for vehicles is by using digital twin technology, which can be utilized for an entire vehicle, its software, electrical system, SoCs, etc., to detect performance issues, test new features and optimize features throughout the entire design and manufacturing processes. Digital twins – virtual models or representations of a system under development – can also be used for vehicle electrical/electronic (E/E)-system validation, allowing for a shift-left approach in automotive design that is invaluable for shortening design and development cycles. That’s because it permits early hardware and software integration and frontload testing. Additionally, silicon lifecycle management technology, with data collected from in-vehicle sensors for analysis, can deliver insights that will support root-cause analysis, predictive maintenance, and aging and degradation management.

1. Centralized Zonal Architecture Will Begin to Come Online

While zonal architectures for automotive won’t be fully rolled out until the end of this decade, we expect to see the development of zonal architectures that will further integrate functions and associated application software into centralized SoCs and modules. This mixture of new applications and evolving system architecture has enabled a new level of sensors and SoCs with high performance, expanded functionality, new hosted applications and amplified amounts of AI. Thus, this increasing level of complexity is forcing the industry to adopt leading-edge automotive semiconductor technologies at 7 nm, 5 nm and even 3 nm. Full realization of zonal architecture within the automotive industry will occur incrementally as certain OEMs, especially electric vehicle manufacturers such as Tesla, invest more aggressively in new platforms and applications and therefore move SoCs to more advanced nodes. In 2024, we also will see more hybrid architectures being used to achieve a centralized system by 2030.

1. New Entrants to Automotive Silicon Design Call for Consulting Services

Much like tech behemoths Meta, Apple and Amazon have brought chip development in-house, we are seeing a similar shift in the automotive industry. Both traditional automakers and those that are relatively new to the scene such as Rivian, BYD Auto and XPeng are getting into the chip design game, but they can’t do it alone. These companies will need their design engineering teams to ramp up productivity quickly. With today’s market pressures and distributed design teams, anything that brings greater efficiency could be turned into a competitive advantage.

A common platform of chip design tools simplifies the effort and fosters better outcomes, eliminating the need to spend time making the tools work together. Beyond tools, additional expertise and consulting will be necessary for these companies, especially given the unique knowledge and technology required to develop automotive-grade products. There are simply not enough chip design engineers with automotive-grade expertise at each of these new-entrant automotive companies to round out an entire department without the use of consultants.

1. UN Regulation 155 Will Increase Pressure for Cybersecurity

Beginning July 2024, the UN 155 regulation on cybersecurity will require the entire automotive value chain from OEMs to suppliers and sub-suppliers to implement a cybersecurity management system. While the industry already can reference and utilize the ISO/SAE 21434 standard, which provides the framework to develop a secure product in the automotive industry, it is a standard, not a legal requirement. UN 155R is a legally binding regulation that automakers must follow to participate in the EU. In other words, the ISO 21434 industry standard provides support to meet requirements found in WP.29 regulation.

This new standard applies to software, systems, components and IP, and will require further collaboration throughout the entire automotive supply chain. There are various ways to prepare now for UN 155, including implementing a Threat Analysis and Risk Assessment (TARA) program, utilizing chip design and IP solutions developed with security in mind, using static analysis solutions to detect vulnerabilities early on, and implementing well-known encryption standards, to name a few.
We will see some major changes in the automotive industry next year. The industry will need to come together to deliver on upcoming regulations, new features that the consumer has come to expect, and faster times to market.

What Lies Ahead for Auto Industry in 2024

Trends in 2024 will demonstrate the need for extended collaboration at all stages in the automotive development process.

www.wardsauto.com

 

[Daniel Nenni]

I can confirm this is true in California!

Tesla drivers had highest accident rate, BMW drivers highest DUI rate, study finds

Tesla drivers had 24 accidents per 1,000 drivers during the period from mid-November 2022 to mid-November 2023, according to a study by Lending Tree.

www.cnbc.com

 

[Xebec]

This article hints at the 'danger' of the Chinese automakers vs. legacy OEMs.. but misses the obvious lack of integrated software and hardware focus that the Chinese (and Tesla) have that legacy does not have. It's not about moving 5-7 year refresh cycles to 3 years, it's about having a cohesive software strategy, and not outsourcing so much that you have no engineering control. Perhaps the 'zonal architecture' is the buzz word Ward's is using, but it seems like Tesla (and the Chinese?) are already doing this in 2023, not 2030..

2024 is also the year we'll start to really see which legacy OEMs properly invested in the battery/electric motor supply chain as electric car sales continue to take more market share.

 

[Xebec]

I can confirm this is true in California!

Tesla drivers had highest accident rate, BMW drivers highest DUI rate, study finds

Tesla drivers had 24 accidents per 1,000 drivers during the period from mid-November 2022 to mid-November 2023, according to a study by Lending Tree.

www.cnbc.com

FWIW - This article is making it's rounds and appears to be based on people seeking insurance on new cars who were recently in accidents with other cars:

No, Tesla, Subaru, and Ram drivers don't have the most accidents

Recent claims that never should have been spread, now fully debunked

brandonpaddock.substack.com

You're correct that we can't know if the vehicle they're looking to get insurance for is the vehicle in which they had the driving incident. Ultimately, the people looking for auto insurance for a Tesla using QuoteWizard from Nov. 14, 2022, to Nov. 14, 2023, had the highest percentage of accident incidents. We made no claims about the safety of Teslas and/or the safety features of any other brand mentioned. Our study is solely about QuoteWizard users during the previously mentioned time frame, the brand of car they're looking to insure, and the driving incidents of those users.

 

[cliff]

Eliminate the tired truck drivers.. makes sense.
Eliminate pizza deliverers... love it!
Eliminate the amazon delivery people.. perhaps

But for transporting people? Is this mainly for transporting illegal aliens that don't have valid drivers licenses?

 

[blueone]

A Cadence website post explaining zonal architecture.

What Is Zonal Architecture? And Why Is it Upending the Automotive Supply Chain?

<a href="/cadence_blogs_8/b/breakfast-bytes"></a> Earlier this week I wrote about the electric vehicle (EV) transition, how it is happening faster than many companies are prepared for, and the implications for the massive amount of change that the automotive companies have to accomplish (see my...

community.cadence.com

Personally, I'm not even a little convinced that the legacy vehicle manufacturers in the US, Europe, and Japan represent the future of... what shall we call it?... personal mobility. They all remind me of, in my field of datacenter computing, the 20th century mainframe vendors. Sweeping change generally requires visionary thinking, and I can't see any evidence of visionary thinking among the current ICE manufacturers.

 

[hkwint]

When it comes to visionary thinking, it's not what's needed for a switch to EV's. There is no rocket science involved at all.

Apart from the "semiconductor" issues mentioned, which I do appreciate, but there are much bigger problems when it comes to the switch to EV's.
I think the mentioned semiconductor-developments are details and are pretty much moot, in the bigger picture.

Nobody at this moment is "not buying or postponing to buy" a car, because the software or computer chips are not good enough.

What is however stopping or postponing people to buy a car:

-Many people in the 'developed' world have been spoiled little princes and princesses, used to be able to buy stuff with money they _don't_ have. That's becoming more expensive however.
-Car makers have to invest a lot of money into the transition, but that's also becoming more expensive; and sometimes ROI comes too late; financing the transition is under pressure.
-Lack of electro engineers to put up EV charging stations / charging points at home,
-Lack of maintenance plan, a lot of EV charging points seem to be broken,
-Lack of a plan for the electricity network, many electricity networks are probably not good enough if 50% of the current ICE drivers want to switch to EV.

There is not much "visionary" thinking involved to solve these issues; a 7 year old can solve these:
- people and companies need to save before they can spend, and
- the electricity network needs more labour, and there need to be maintenance plans.

Early adopters already adopted an EV, mostly richer people, and (at least where I live) the richest people receive the most subsidy to buy expensive electric cars.
So mainly the poor are paying the rich to transition to EV's.
The rich transitioned, are buying solar cells to charge their EV's, while the "poorer" have to pay; everyone starts to notice now and subsidies are halted.

Also, looking at the electricity grid:

-Even a country with a pretty well developed electricity grid, such as "EV leader" Norway with >20% of the people driving electric*, already has an electricity grid crises;

The Norwegian Power Adequacy Crisis is Accelerated by Intermittent Renewables Exploiting Grid Reserves

Solar and wind still do not pay their fair share in system costs, grid inertia, and expensive backup power. It is a threat to the security of our electricity supply.

jonasnoeland.substack.com

-The Netherlands, with 4,5% of the people driving electric, with a less developed power grid; mainly due to gross lack of a government vision the last 20 years, already has an electricity grid crises:

Netcongestie de problemen van een overbelast stroomnet

In vrijwel heel Nederland is er momenteel sprake van netcongestie. Dit leidt tot allerlei problemen op het elektriciteitsnetwerk, variërend van geen mogelijkheid voor nieuwe aansluitingen tot tijdelij...

www.archidat.nl

-In a developing country as the US, which doesn't even have a national power grid AFAIK and the regional grids can almost break at any point, and only 1,3% of the people driving electric, stuff is going to break if more people want EV's.

https://www.washingtonpost.com/business/2023/11/08/power-grid-blackouts-texas/

Two-thirds of the U.S. is at risk of power outages this summer–but it’s not stopping Americans from electrifying everything in their homes

Americans' concerns over the power grid's reliability and a growing electrification movement are fueling solar installs.

fortune.com

Then there are the even less fortunate countries, such as India, Bangladesh and Pakistan, for them it would be even harder to transition to EV's.

The problem is, building "new shiny things" is always more favored by politicians, than "maintaining things".
Which is the reason why the US, Germany and The Netherlands have so many outdated road bridges on the verge of collapse, and why in Italy a bridge even collapsed in 2018.

The solution is simple; train more people as electricians, fund new high voltage power lines, and actually maintain it.
Nothing visionary needed!

*https://en.wikipedia.org/wiki/Elect...use_Top_countries_&_regional_markets_2020.png

On the software side, people are reluctant to buy a new car, one of the reasons being because in 3 years the software / chips will be outdated.

I'm lucky to own a "quite old" 2011 car, which has a lighter-slot and 3,5mm line in. You can buy a bluetooth receiver for € 7, connect one end to the 3,5mm jack for sound and one end to the lighter slot for power, and voila, your 2011 car has bluetooth!

However, let's say you have a 2018 car which has USB-A, and an "old" version of bluetooth, and probably outdated maps, and an outdated version of Android, how can you update it if you still want to use it in 2028?

For the electronics, it would be great to have components which you can swap. Like the old DIN-radio format, but then for car electronics and entertainment.
However, it seems auto makers are more and more opting for monolithic systems instead of modular things.

Another example, batteries: If you buy an old 2015 Nissan Leaf, with short range, you can upgrade it with bigger batteries and "quick charge", and more watts to charge.
But the price of retrofitting / upgrading will probably be close to the cost of buying the newest Leaf, which then has the same range.

Which means, these old EV's are almost worthless, and now that seems to be one of the reasons people are afraid their EV will have low resale value.

This is also a place, where "modularity" can improve things. It should be easy to swap out charging sockets or batteries, if in the future newer and better battery / charging technologies arise. That way, with modularity, cars keep their resale value.

If tires were not replaceable and a monolithic part of a car, almost all cars would be scrapped within 5 years, and many people would not buy one because resale value is too low.
"Modularity" of car tires solved this.

But of course, auto makers also want to try to make their cars outdated by making monolithic designs, so right now what the consumer asks is different to what the auto makers are offering.

 

[blueone]

Wow, what a post, but the concept I'm referring to is modularity of design, especially for computing and sensor resources, as you alluded to with DIN chassis for radios. The innovation that's necessary, in my opinion, is in the way vehicles are designed and manufactured. Even Tesla is just a bit ahead of the legacy manufacturers in this regard, what with more sophisticated and centralized computing and over the air software upgrades, but even a Tesla still looks to me like it has a pretty inefficient legacy design model.

I completely disagree with your assertion that visionary thinking isn't required.

First of all, for EVs the notion that there's a monolithic design that's improved every 4-year generation should be obsolete. Electric motors are already very efficient, Tesla claims 97% on the Model Y, so the usual improvement in motors is probably irrelevant. Perhaps motors can get physically smaller and lighter in weight, or longer lasting, but from what I've read the improvements will not be drastic. Things like all-wheel drive are more a matter of computing in an EV than the mechanical nightmare on an ICE vehicle. EVs are already quicker and faster than most sane people need for street driving. Making them quicker or faster is irrelevant. Suspensions are already very sophisticated (active adjustment and magnetic dampers). Batteries are a chemistry problem and will improve at a pace different from electronics, and I agree with you they should be modular, but batteries weigh so much and take up so much space I suspect the fully modular solution is farther off than anyone wants it to be.

I think the innovation the industry needs is to view an EV as just a platform that consists of a bunch of upgradable and interchangeable modules, especially in computing and sensors, and then every 5-7 years (just a guess) the vehicles could get refurb'd with modular computing components, just the way rack servers in a data center are. You don't buy a vehicle as a unit, you buy a subscription to the technology improvements, hardware or software. The entire notion of proprietary, each model custom-designed from the ground up for a specific generation needs to go out the window, because when one company does that I don't think the ones that don't will be able to compete. Standard form factors are all over computing, I think they need to be all over vehicles too. And I doubt the legacy vehicle builders will seriously consider this notion.

 

[Paul2]

Digital twin technology can shorten automotive development cycles.

Digital twin will increase automotive development cycle

Automakers are rushing to keep up with Chinese manufacturers who are setting the pace of development at levels previously unseen in the traditional market. Due to robust automotive safety standards and resulting testing needs, the technology in a brand-new car is likely to be five to seven years old. Chinese manufacturers are closing that gap to be closer to three to five years. Both silicon and software testing validation need to happen quickly to accelerate development, often in parallel with one another.

Safety critical chips running ABS, brake control, steer-by-wire, ESP haven't changed in many, many years, and those which did likely got cheaper, and more integrated. 9 out of 10 of car makers today just receive a Bosch black box.

Nobody cares about a video player in a car hanging up, because it is separated from dumb electronics running driving aids, or let alone ABS (Tesla being one big exception).

Digital twins – virtual models or representations of a system under development – can also be used for vehicle electrical/electronic (E/E)-system validation, allowing for a shift-left approach in automotive design

Speciality engineering companies take off that task from everyone, sans WV, and Toyota, and they have long perfected the process without using anything like that. The non-infotainment connected part of car electronics can be designed by them literally within weeks, and pass the first live test. It's very well working copy & paste type of work, and I see no desire to disrupt something so critical, and dangerous, and working well just for marketing influenced takes.

1. Centralized Zonal Architecture Will Begin to Come Online

I already see a "less-digital" architecture coming online. Excesses like 1 microcontroller per-button are gone on some of the most expensive cars coming around next year, because people learned: 1 LED, switch, and a resistor work well, and locking up production of a $100k car because of one microchip which just reads 1 button doesn't work well.

Much less chips, and much more direct wiring Less busses, and much more "hard" system separation (just basic GPIO signalling for part status.)

 

[Xebec]

Nobody cares about a video player in a car hanging up, because it is separated from dumb electronics running driving aids, or let alone ABS (Tesla being one big exception).

They're not connected on a Tesla. You can reboot the car computer and autopilot, auto braking, and safety features all still work.

 

[Paul2]

They're not connected on a Tesla. You can reboot the car computer and autopilot, auto braking, and safety features all still work.

They are https://teslamotorsclub.com/tmc/thr...oblems-we-need-nhtsa-to-force-the-fix.216615/

The media player failure on Teslas will not crash a moving car, but it will trigger a failsafe, and prevent the car from starting. Effectively, Teslas are immobilised by a media player failure.

 

[blueone]

Dan Neil of the Wall Street Journal reports 3D printing could be part of the automotive future. (Sorry, WSJ is pay only.)

Remaking Automaking for the Future

Bespoke personal transportation is within reach. To get there, carmakers will need to change how they design, build and sell cars.

www.wsj.com

IMO, not practical for ICE vehicles or pickup trucks and heavy duty vehicles.

 

[Xebec]

They are https://teslamotorsclub.com/tmc/thr...oblems-we-need-nhtsa-to-force-the-fix.216615/

The media player failure on Teslas will not crash a moving car, but it will trigger a failsafe, and prevent the car from starting. Effectively, Teslas are immobilised by a media player failure.

Hmm I’ll look into this a bit more. That post is referring to the first generation MCU on Teslas when they were a low volume manufacturer.

I know MCU2 and MCU3 are not affected by the eMMC recall, and on my Tesla (MCU2) I can reboot the computer while driving without autopilot or any safety features ‘stopping’. The only thing you lose driving wise are lack of speedometer, and no turn signal sounds since they’re created by the computer and played through your sound system.

Thanks for the info.

 

[tooLongInEDA]

Wow, what a post, but the concept I'm referring to is modularity of design, especially for computing and sensor resources, as you alluded to with DIN chassis for radios. The innovation that's necessary, in my opinion, is in the way vehicles are designed and manufactured. Even Tesla is just a bit ahead of the legacy manufacturers in this regard, what with more sophisticated and centralized computing and over the air software upgrades, but even a Tesla still looks to me like it has a pretty inefficient legacy design model.

I completely disagree with your assertion that visionary thinking isn't required.

First of all, for EVs the notion that there's a monolithic design that's improved every 4-year generation should be obsolete. Electric motors are already very efficient, Tesla claims 97% on the Model Y, so the usual improvement in motors is probably irrelevant. Perhaps motors can get physically smaller and lighter in weight, or longer lasting, but from what I've read the improvements will not be drastic. Things like all-wheel drive are more a matter of computing in an EV than the mechanical nightmare on an ICE vehicle. EVs are already quicker and faster than most sane people need for street driving. Making them quicker or faster is irrelevant. Suspensions are already very sophisticated (active adjustment and magnetic dampers). Batteries are a chemistry problem and will improve at a pace different from electronics, and I agree with you they should be modular, but batteries weigh so much and take up so much space I suspect the fully modular solution is farther off than anyone wants it to be.

I think the innovation the industry needs is to view an EV as just a platform that consists of a bunch of upgradable and interchangeable modules, especially in computing and sensors, and then every 5-7 years (just a guess) the vehicles could get refurb'd with modular computing components, just the way rack servers in a data center are. You don't buy a vehicle as a unit, you buy a subscription to the technology improvements, hardware or software. The entire notion of proprietary, each model custom-designed from the ground up for a specific generation needs to go out the window, because when one company does that I don't think the ones that don't will be able to compete. Standard form factors are all over computing, I think they need to be all over vehicles too. And I doubt the legacy vehicle builders will seriously consider this notion.

Yes, these modular car systems where you can retrofit parts from third party suppliers to progressively upgrade what you originally bought would be great. But please show me any sign that this is actually what is happening. I think the reverse is occurring- and quite rapidly. Companies like Tesla and Apple get very rich by building proprietary closed systems. Just imagine the impact on Apple's bottom line if you could upgrade iPhone storage with market priced memory. The same goes with cars and even craziness like BMW's move to make heated seats a subscription (certainly in some markets and for some models).

This is precisely why - like @Barnsley - I prefer older tech cars that don't cost a fortune to repair when some non-essential piece of tech that I never need or use goes wrong (this is tech and 'safety' stuff that is deemed essential by governments/the EU after lobbying by car companies bent on pursuing planned obsolescence).

It's the same reason I have old steel-framed road bikes. They are easy to fix and upgrade. At some point a bike, car or computer becomes "good enough" and there's something better to spend the money on.

I struggle to see any way that we can force car manufacturers to produce modular vehicles (actually complex systems these days). However much we as consumers want and need that. The best we can hope for is punitive taxes on car companies for cars which are scrapped after - let's say - less than 10 years without any good reason. Early scrappage is a massive waste of resources and energy.

 

[blueone]

Yes, these modular car systems where you can retrofit parts from third party suppliers to progressively upgrade what you originally bought would be great. But please show me any sign that this is actually what is happening. I think the reverse is occurring- and quite rapidly. Companies like Tesla and Apple get very rich by building proprietary closed systems. Just imagine the impact on Apple's bottom line if you could upgrade iPhone storage with market priced memory. The same goes with cars and even craziness like BMW's move to make heated seats a subscription (certainly in some markets and for some models).

I can't show you any sign modular design is happening, and I agree for the short term proprietary is making a last stand.

As for upgrading iPhone memory, I doubt that would be possible even if Apple wanted to allow it. Apple puts LPDDR5 DRAM in a separate IC package that's directly on top of the CPU SoC, and they use TSMC InFO-PoP, according to industry reports.

This is precisely why - like @Barnsley - I prefer older tech cars that don't cost a fortune to repair when some non-essential piece of tech that I never need or use goes wrong (this is tech and 'safety' stuff that is deemed essential by governments/the EU after lobbying by car companies bent on pursuing planned obsolescence).

My spouse is another dinosaur like you two. She likes analog everything, even a manual transmission. Your concerns about longevity are justified, IMO. Will a replacement 15" LCD instrument panel for a BMW be available 20 years from now. I'd bet against it.

It's the same reason I have old steel-framed road bikes. They are easy to fix and upgrade. At some point a bike, car or computer becomes "good enough" and there's something better to spend the money on.

Even my analog spouse has a carbon fiber bike, so you're on your own with the steel tube thing.

I struggle to see any way that we can force car manufacturers to produce modular vehicles (actually complex systems these days). However much we as consumers want and need that. The best we can hope for is punitive taxes on car companies for cars which are scrapped after - let's say - less than 10 years without any good reason. Early scrappage is a massive waste of resources and energy.

As I mentioned in Post #6 of this thread, I have no confidence that the legacy auto manufacturers in any country will willingly go modular. And I don't think the government can successfully mandate it, nor would they have any idea how to do it. The only force strong enough to budge them will be a competitive company that brings a new, lower cost and more agile strategy to the vehicle market. Remember mainframes from the 1960s-1980s? All completely proprietary from their cabinets to their electronics to their software to their networking, and even their external storage. And then microprocessor-based standard high volume rack-mounted servers came along running UNIX, Windows Server, and Linux, and there's only one mainframe maker left, serving the computing equivalent of the super-luxury market. I think a similar sort of strategy can be used for EVs, and that it may take several years to appear, but all it will take is one car company which figures out how to undersell the competition while making a higher gross margin than they do.

 

[Tanj]

Vehicles already are very modular. Companies like GM or Ford already are mostly final assembly of subsystems from outside suppliers. VW and Toyota have maybe more captive subsystems but they still build a modular product, it is necessary to be competitive in having a variety of models to segment and saturate the consumer space.

What they do not have, including new entrants like Tesla, is any incentive to make that modularity accessible to the customer. Heck Tesla sets the pace here by turning things off when you sell a vehicle, the new owner needs to pay to have soft features reactivated. So there is no sign the EV vs ICE is becoming more consumer friendly. Sure, an old ICE maybe repairable and hackable, but the brands understand the "freemium" revenue model involved in capturing all upgrades and maintenance, on all new vehicles.

Heck, even bikes. Try adding new features to that built-in trip computer without buying them from the maker. Or hacking the pedal-assist behavior.

This is not a technology problem. It is the business model.

 

[blueone]

Vehicles already are very modular. Companies like GM or Ford already are mostly final assembly of subsystems from outside suppliers.

The current modules are mostly proprietary to a particular vehicle manufacturer. You can't even share these modules between vehicles from the same manufacturer, like VW, Audi, and Porsche.

VW and Toyota have maybe more captive subsystems but they still build a modular product, it is necessary to be competitive in having a variety of models to segment and saturate the consumer space.

VW depends a lot on Bosch. BMW uses a lot of Harman electronics, but, again, just specials done for BMW.

What they do not have, including new entrants like Tesla, is any incentive to make that modularity accessible to the customer.

Agreed, but I'm not really talking customer accessibility. I'm talking about upgradability, which requires standard interfaces, for hardware and software.

Heck Tesla sets the pace here by turning things off when you sell a vehicle, the new owner needs to pay to have soft features reactivated. So there is no sign the EV vs ICE is becoming more consumer friendly.

I didn't mean to imply that EVs are currently any better. My intent was only to point out that with EVs many mechanical differentiators go away.

Sure, an old ICE maybe repairable and hackable, but the brands understand the "freemium" revenue model involved in capturing all upgrades and maintenance, on all new vehicles.

Yup, until they can't be cost competitive that way.

It is the business model.

That's the truth. We'd still be using proprietary computers if the legacy computer companies had their way.

 

[Tanj]

Yup, until they can't be cost competitive that way.

In a sales world, freemium wins because an upstart creating a "forever" product cannot win on the initial sale. But in a lease world "forever" may have enough advantage if the bet is on holding resale value.

 

[hkwint]

I think the innovation the industry needs is to view an EV as just a platform that consists of a bunch of upgradable and interchangeable modules, especially in computing and sensors, and then every 5-7 years (just a guess) the vehicles could get refurb'd with modular computing components, just the way rack servers in a data center are. You don't buy a vehicle as a unit, you buy a subscription to the technology improvements, hardware or software. The entire notion of proprietary, each model custom-designed from the ground up for a specific generation needs to go out the window, because when one company does that I don't think the ones that don't will be able to compete. Standard form factors are all over computing, I think they need to be all over vehicles too. And I doubt the legacy vehicle builders will seriously consider this notion.

Well, that was exactly my argument )

Visionary thinking is mostly required when it comes to cost. Just look at Tesla lowering the price of their cars, and BYD overtaking them as biggest supplier of EV's.
And also when you talk about electro motors, the biggest improvement needed is cost.

Petrol / Gas prices declined, EV subsidies are diminished (everywhere except the US), so the car industry is postponing EV investments in this financial climate where money isn't free anymore.

 

[hkwint]

...

As I mentioned in Post #6 of this thread, I have no confidence that the legacy auto manufacturers in any country will willingly go modular. And I don't think the government can successfully mandate it, nor would they have any idea how to do it. The only force strong enough to budge them will be a competitive company that brings a new, lower cost and more agile strategy to the vehicle market. Remember mainframes from the 1960s-1980s?

Well, when it comes to phones, "upgradeability" is not a big issue, because "resale value" is not a marketing point, for smartphones.

But suppose you are not (considered) a premium brand, let's say you're BYD.
When it comes to cars, "resale value" could be a good marketing point. You just have to inform your customers.

When it came to the "right to repair" or finally getting rid of the proprietary Apple only Lightning standard, it was basically the EU pushing this agenda. Because the EU doesn't produce smartphones, that was a non brainer to do.

But when it comes to cars, I don't see the EU (nor the US, China or Japan) pushing for "right to repair" your car battery pack, or aiming for 1 socket for EV's. So any push for modular standards would have to come from consumers.

Which is probably the reason why there exists interface standards for hyperscale datacenters: Because companies like Facebook actually demanded it.