Automotive is broken or Why cost is running up but your vehicle system is not

Once envisioned as the software powerhouse of Europe’s biggest carmaker, its dedicated software unit is now refocusing on coordination rather than creation — a quiet signal of the industry’s struggle to reinvent itself for the software-defined era.¹ European automotive OEMs and suppliers are reducing workforce in the tens of thousands, some using layoffs, some still claiming to just using natural attrition. But the writing’s on the wall: Many OEMs, especially the German premium segment within Europe’s automotive industry is coming down severely.

Let’s dive into some aspects why.

Abstract

A central promise of Software-Defined Vehicles (SDV) is the evolution from a car based primarily on mechanics or hardware to a car that is primarily electronically controlled and therefore primarily dependent on software.² Yet, all players in the automotive industry face the big C challenges: Complexity, Chain of supply, Cost pressure, Climate.

In this essay, we will first clarify these challenges: Who creates the cost pressure and why? What is the effect of the cost pressure? What is the usual organizational structure in automotive industry? Why is the chain of supply a challenge, and what specifically is the complexity challenge?

Then, to clarify the effect on the individual companies and organizations, we will have questions and discussion prompts that can serve as guidance to particular solutions. You will gain an understanding of some of the systematic issues that your organizations may need to address.

Furthermore, I will describe a potential way forward through the multitude of crises, towards regaining strength for innovation by great engineering for electronics in vehicles.

Complexity

What does complexity even mean in this context? And why is that a challenge?

Gimmicky features like oversized screens or even touch screens, like a full width moulded screen. Makes the car expensive, does not have real utility. Pure luxury, but waste of money for all, ties up resources for developing. It does not bring you any faster from A to B, does not make your journey more safe but au contraîre: It poses an additional risk of distraction without level 4 automated driving, which is not happening any time soon around here, and it increases any cost of damage dramatically.

Centralization and consolidation of ECUs: good intent: Centralization reduced wiring and improved bandwidth utilization, but also concentrated functional and safety dependencies in a few massive systems.³ As a result, every update now affects more interfaces, and overall complexity grows instead of shrinking.⁴ Those who argued that a central ECU is just like an IT server on wheels have never maintained an IT server, or never put an ECU on the road, or both.

To be fair: the electronic age and the software age have brought a lot of real advancement in automotive. A lot of the important safety related innovations have already been made: ABS, ESP, active steering, emergency break assistant. ADAS is somehow on the rise with multiple assistants, as, e.g., lane assistant, cruise control, and even a traffic jam assistant. The latter being an irony in itself, because „you are not stuck in traffic jam, you are the traffic jam“, so a function in an automobile provides a remedy to a problem created by that very automobile’s existence.

On the other hand, a lot of software related so-called innovations are a real bummer. Speech recognition: still sucks, rather than providing a distraction free command entry, drivers still have to pay full attention to how exactly and what exactly they speak, and carefully listen to whatever misunderstanding the system responds. Touch screens: distract from street attention compared to the knobs and dials from the past. Infotainment systems: gimmicky. Navigation: got better on a functional level, but placed a full fledged 3D simulation into the central screen for no reason.

Automated driving (AD) is kind of dead, or at least comatose. A famous south German OEM had built a new campus in the north of Munich for thousands of software developers for automated or even autonomous driving, but this campus has been ramped down and is mostly deserted of automotive software engineers now.

Cars are still growing bigger and fatter, with demands to even make parking stalls wider to fit the obese vehicles.⁵ Exterior design is no longer a virtue that car makers embrace,⁶ but something that tends to look like you handed a three year old some crayons and too little time.

The true cost of complexity lies not in technology, but in misaligned incentives.

Chain of supply

A great push in the last years has come from initiatives to separate hardware from software. Formerly tightly integrated systems shall, according to OEMs, be broken up into hardware from one supplier and software from other suppliers, and they talk about better layering, managing complexity and such. Separation of HW/SW is the classical divide et impera, but not just for technical impera, but also to break up strength of Tier1 suppliers and sourcing HW and SW separately, and easier build up second and third source.

From a purchasing standpoint, it seems to make sense, because the very existence of the term Tier 1 supplier indicates there are Tier 2 and Tier 3 suppliers. Now, modern economics teach us, OEMs seek to cut the middle man, and Tier 1 is seen as exactly that.⁷

OEMs therefore, not all, but at large, had the grandiose idea to aim to in-source software and E/E competency, mostly by creating large spin-offs for E/E and software. So in-sourcing then out-sourcing, but nearer. Those are separate legal entities, with all the effects of now having to officially purchase stuff from what previously had been just a department or business unit. By separating these new entities legally and financially, OEMs created powerful incentives for internal billing and scope negotiation rather than for technical coherence. Consequently, the very structure intended to build autonomy made integration harder — leading to the well-known failures of these software organizations.

It gets even better: from day one, these newly created or cut-out organizations needed personnel. A lot of personnel. Thousands. Fortunately, the OEMs found those people: the experts that formerly worked with Tier 1 suppliers as Steuergerätesachbearbeiter, as ECU clerks. Over night they became software experts. This is incredibly cruel to push this onto people with little or no additional training. It is a very different thing to be a builder vs. having to architect and laying the bricks yourself. A very, very different thing. But the hiring machine geared up, and drained the market from software experts, both automotive experienced and from other fields, since “we think automotive software in a new way” (intentionally unattributed but common quote).

It turned out that the laws of electrics, of electronics, of power use and computer science in embedded still applied, and we’ve seen fabulous failures of those organizations, shifting SoPs, bugs galore, non-functioning software, and as a result eject chairs on management level, and job crisis on workforce level. It turns out that even the spun-off software organizations did not exactly know what they were, or what they should be, but understood that somehow it’s now their fault. So, as result, we do see a strong decline in those organizations’ scope.⁸⁹

Because internal spin-offs lacked mature delivery processes and supplier networks, OEMs had to return to their Tier 1 partners. This U-turn consumed precious lead time, increasing schedule pressure just as software release cycles shortened.

The effect is now we have OEMs, their software organizations, and still Tier 1 suppliers, so more players than before, and therefore a complex projects even just organizationally, not even talking tech.

While organizations and technology have evolved, the ways of purchasing systems, subsystems or engineering have not evolved as much, leading to OEMs and suppliers lurking each other, always on guard not to be squeezed out any further. Yet, many companies do so with little lever, especially suppliers with their often lower relative power compared to large OEMs.

Often times, on working level, engineers across organizations just want to create great stuff and get compensated for it. On top level, leaders often speak about partnership, but it remains unclear what that means.

Cost Pressure

Automotive companies regularly employ high 4 digit to 5 digit numbers of engineers, factory workers, administrative staff. Developing and producing cars, after all, is labor intensive and thus expensive if doing it on your own as a company.

Companies of all times have tried to bring down labor cost, by outsourcing to cheaper local companies, then to other countries, then to other continents. The whole idea of the open market is for companies in various countries to specialize in their production or services and be better and cheaper in offering those. Some vehicle parts are produced regional, other parts are shipped across the oceans. The same happens with development labor, just that the results of that development work is much easier to ship digitally.

Of course, since all automotive companies are pursuing an ever lower price per development and per piece, all are looking for all possible ways to achieve that. And yes, that also means over time escaping countries with strong unions and high cost. At least when looking at a vehicle development and production as a relatively fixed and given process and known thing. And yet, Chinese and Indian car companies not just still have way lower personnel cost than European ones, but find fresh, clever ways to do things differently and, after some years of learning and failing, develop vehicles and vehicle functions faster. Already in all cases, with all aspects of homologation covered in first shot? Of course not, but learning at fast pace.

With all the separation of hardware and software thing going on, of course the former piece price of the integrated system now gets split up into two components. Of which, from purchasing standpoint, only the hardware one as a weight and thus a part number and so according to classic manufacturing has some value attached to it. The software, not so much.¹⁰ Because a lot of the software is nondifferentiating, so some low price might be accepted, but mostly nowhere near the cost of software development.¹¹ Splitting the hardware and software contracts gave purchasing departments more negotiating levers — which created a race among suppliers to underbid each other on partial scopes.

This dynamic, meant to cut cost, instead increased total system cost as integration effort exploded. It turns out the software/hardware separation was just one more means of cutting cost and using competition among supplier to bring everyone down. Yet later OEMS have to spend increasing amounts of money on engineering service providers that shall duct tape everything back together to get the vehicle moving at all.¹²

Climate

Overall, how did we get here as an automotive industry, and especially in Europe, and in Germany in particular?

Well, it looks like the OEMs lobbied the shit out of the tax system to prevent having to think about a changing mobility and business landscape and changing society. And now, seemingly suddenly, the overall situation has changed. New–well in fact not so new but long ignored–players emerge, e.g., from China, former purely domestic brands push into Europe, and from the tariff policies it really does look like a battle at times. China, lobbyists scream, they are subsidizing their automotive industry to make it stronger internationally. Doesn’t sound familiar? Dienstwagenprivileg, anyone?

So, effectively a significant share of the German OEMs’ business model is to sell vehicles as company cars, where the price sensitivity is not as dense as when selling to private end users.¹³ While some businesses’ employees really need a vehicle to conduct their value creation, most person vehicles that run as company cars serve as non-salary employee benefit for highly paid white collar workers.¹⁴

I may be exaggerating a bit, sure, but it’s basically that. Especially the premium segment of OEMs domestically covers mostly the business model as laid out in the previous paragraph: It would not be there in the way it is without the excessive subsidization.

Effectively, of the 13,7 billion Euros directly and indirectly going into company cars as state subsidies, we can estimate that half of that is effectively spent on benefits for people that would do well without, too.¹⁵

And despite the massive success of this business model in the past decades, OEMs still face tremendous price pressure, because the niche of customers being able to afford luxury cars out of their very own pockets is finite. Backing out of the regular low to middle class income segment as bread and butter business may cater for excellent margin per vehicle, but lower overall number of sellable vehicles.

Now the problem is not just large amounts of citizen-funded tax money spent on corporate subsidies, it is the larger effect of those, too.

Premium cars tend to have larger combustion engines. If gasoline is free for the white collar employee, and subsidized for the company, then why not pour some liters more through the engine in order to get better acceleration, growling sound, and the feeling of social status?

A BMW engineer once put it bluntly: “We’re not selling a more complex machine — just a bigger bore in an engine that’s already been perfected.” It was a quiet acknowledgment that progress doesn’t always mean adding more parts. Sometimes, it’s about how much value the market perceives in what’s already there.”

That party was nice as long as it lasted, but now it’s close to its end. Unfortunately, the industry lobbied a long time to keep the business model in place, so well that meanwhile even with OEMs’ executive management beginning to embrace electrification of drive trains, politics is trying hard to maintain the status quo and lean against the dying of the combustion engine. By tying industrial policy to short-term job protection, politicians delayed adaptation and weakened incentives for early EV investment. Consequently, when global demand shifted, European OEMs faced the transition with higher cost and less readiness — the textbook definition of backfire. And we are exactly at the beginning of this backfiring, because now in a sudden everything happens or has to happen pretty fast.

Already nine years ago, we were consulting a business unit of an automotive supplier that specialized in electronic control units (ECU) for gearboxes, being tasked to help them with identifying future-proof system and software architectures. Long story short: Turns out in a purely electric drivetrain there is no gearbox anymore, and thus no need for gearbox ECUs, and thus the whole business model of that unit was bound to become void. You can guess that changed the characteristic and scope of that consulting engagement.

The whole business becoming void is happening to a lot of suppliers that deliver components for combustion drivetrains, and even for hybrid electric drivetrains, and also for manufacturing plants of OEMs as well as the very development departments of those OEMs whose main contents of work will vanish or has already begun to vanish.

Because the internal combustion engine (ICE), despite being a really magnificient fine piece of engineering, with fully perfected details in all aspects, and decades over decades of technical brilliance going into it, was always a bridge technology to somehow make a vehicle able to make use of an onboard energy source. Remember, electric cars have been there before gasoline powered cars.¹⁶

The electric car, of course, does neither solve the problem of congested streets, nor the one of excessive resource consumption for production, nor excessive surface sealing for roads and dehumanizing cities.¹⁷ But they do partially solve the issue of local air pollution and CO2 production.

Other countries have embraced this much faster than central Europe. Where China is already at near 50% of new vehicle sales being battery electric and that percentage still rising, Germany has fallen back again in both absolute and relative numbers to below 20% in 2024. ¹⁸ It’s not all of Europe: Netherlands is on a steady raise almost like China, and Norway exceeds 90%.

And so, we do see former domestic brands of China pushing hard into the European market with aggressive pricing, since our local brands still fail to address the lower price market adequately. This situation was created over decades of policy-making and lobbying, and protectionism will only delay yet amplify the dramatic effects.

Questions to ask yourselves

There is no going back in time. The old times are not coming back. No matter how strong you pull ties into politics, times are a-changing. Dealing with software intensive systems will not get easier anymore. Preparing your E/E architecture for supply chain crises will not get easier anymore. Selling cars will not get easier anymore.

I have been working for decades in this industry now, so I know how it feels, and I understand the sentiment and even anger and urge to try molding the political constraints to create a bubble of eternal business condition stability. But that won’t solve the problems, not even on technical level.

So what to do?

Always be curious on all levels. Always be curious on top management level. Always be curious on middle management level. Always be curious on working level. Develop the Always be curious mindset (the ABC mindset¹⁹). Which means, ask a lot of questions yourselves, like the following:

• If company cars would no longer be tax subsidized, would you be able to sell them with the higher end features in quantities that you are able now to?

• What does growth mean, besides financial growth for the company

• With obviously everyone still rushing to capture the premium market because of nicer profit margins, how far is that sustainable for your corporation if you abandon the lower price market?

• How long will your reputation from having built excellent cars in all price ranges last, when most of the population no longer drives your cars? And what is the likelihood of the plummeting reputation to trickle up?

• If you are a famous German brand, how sustainable is success in foreign markets when you more and more lose traction in your home market?

• When you are talking about business partnership, what exactly do you mean, and what does your boss and your organization mean? And how is partnership different from a supplier-customer relationship?

• Why do you prefer speaking of partnership instead of customer-supplier relationship? Why is the one “better” than the other?

• If you had to spent every Euro on your next car from your personal pocket, which features and extras that you were involved in developing would you buy?

It ultimately leads to the essential questions:

What the heck are we doing in our organization and why? And how do our everyday decisions, from technical to product to portfolio decisions, relate to that why?²⁰²¹

The four big C’s — Complexity, Chain of Supply, Cost Pressure, and Climate — are not isolated crises but reinforcing loops of the same system. The pursuit of ever more features adds technical and organizational complexity; this complexity multiplies interfaces across the supply chain; fragmented responsibilities and relentless cost cutting then amplify inefficiency; and all of it unfolds under the growing external pressure to decarbonize and redefine mobility. Each C feeds the others, until innovation itself becomes the victim. To move forward, the industry must learn to treat these not as separate problems to be optimized, but as connected dynamics to be balanced — technologically, economically, and socially.

The way forward

So, what shall be the way forward? I don’t have a crystal ball (although I keep a salt lamp on my desk for when I get questions that would require one), so I have no definite prescription.

Where is the software and technology in that?

You may ask where is the software in all of this? The software is in everything. Embrace the fact that consolidation of ECUs in domain, zonal or central architecture does not make development easier per se,²² it shifts the complicatedness and complexity into different areas, but hundreds or thousands of features will not automatically become less complex to handle just by running them on the same embedded computer with gazillion cores. While you save in wiring and developing distributed embedded systems, you will need to invest in multi-core and multi-processor engineering capabilities.

With all bells and whistles that you add to your products, to your cars, remember the most important aspect is to have system–a car–that actually works, that works reliably. Simplicity is the ultimate sophistication. Yes, you may not be able to argue an ever dramatically increasing price for a model without additional features. Or can you? My stainless steel pan costs three times as much as an average coated pan, and it does not have that non-stick feature. And still it is a much better investment.

And on business organization level?

Accept that all in this industry are in it together.

The main business of car companies, of automotive companies is to allow people to travel from point A to point B, individually, spontaneously. Every other feature of a vehicle is an add-on.

If you are an OEM, you are a mobility provider, which might indeed be a more vague business than selling cars per unit. If you tie yourself to the existing–and old–business model of selling bent sheet metal with electronics included, you may get in trouble sooner or later.

If you are a Tier1 supplier, you are a solution provider, which is indeed a much wider scope than shipping ECU boxes à la carte.

Especially in Europe, and more so in Central Europe, as high cost countries for development and manufacturing, for sure OEMs and suppliers can try to game each other, and for a while that may work, but effectively when you and your company are doing your part to ruin the automotive engineering ecosystem, don’t complain and cry later when the pond no longer feeds you. When everything can be sourced globally, so can end customers “source“ their vehicle provider globally.

On short term, our overall system rewards economic unilateralism, but in the long run thrives on multilateralism, which needs multiple almost equally strong players.

I’ve been working for a company that had one of their software products in strong and direct competition to a similar solution from another company. Over the years and decades–still both products exist–the very fact that those products are in competition highly benefited the performance and technology of both. Neither of the two opposing solutions would be where they are today, had there not been a constantly fruitful competition.²³ My heart is with the thousands of factory workers and engineers that have been and will be laid off in this chain of events. It could have been prevented to a large extent.

Yes, the automotive industry is home of a lot of jobs worldwide, and in Europe, and in Germany. But so are other industries. What if every time we talk about automotive industry we began talking about mobility industry, and actually mean it, to unite with commercial vehicle, train, subway, tram industry?

Again: Accept that all in this industry are in it together.

If you’d like to talk with your organization: do approach me.

References

---

1. Engelking, Niklas Jan. “Cariad: VW-Tochter stellt eigene Software-Entwicklung weitgehend ein.” heise online, October 4, 2025. https://www.heise.de/news/Cariad-VW-Tochter-stellt-eigene-Software-Entwicklung-weitgehend-ein-10711997.html.↩︎

2. Valeo. “Everything You Need to Know about the Software Defined Vehicle (SDV).” December 12, 2024. https://www.valeo.com/en/everything-you-need-to-know-about-the-software-defined-vehicle-sdv/.↩︎

3. Baum, Markus, Konstantin Shirokinskiy, and Shuai Shi. Automotive 2040: Connectivity and Software-Defined Vehicles. Roland Berger, 2024. https://www.rolandberger.com/en/Insights/Publications/Automotive-2040-Connectivity-and-software-defined-vehicles.html.↩︎

4. Broy, Manfred, Ingolf H. Kruger, Alexander Pretschner, and Christian Salzmann. “Engineering Automotive Software.” Proceedings of the IEEE 95, no. 2 (2007): 356–73. https://doi.org/10.1109/JPROC.2006.888386.↩︎

5. “Mehr Raum für rollende Riesen? Sollten Parkplätze immer größer werden?” VCD, March 6, 2023. https://www.vcd.org/artikel/mehr-raum-fuer-rollende-riesen-sollten-parkplaetze-immer-groesser-werden.↩︎

6. Maurer, Marco. “Warum sind unsere Autos so hässlich?” Tages-Anzeiger, April 7, 2025. https://www.tagesanzeiger.ch/tesla-fiat-und-vw-golf-warum-sind-unsere-autos-so-haesslich-723596938656.↩︎

7. Maximilian Majic, Simon Schnurrer, Jörg Giebels, et al. Die Nächste Hürde: Wie Finanzieren Automobilzulieferer Die Transformation? Studie. OliverWyman, VDA, 2024. https://www.vda.de/dam/jcr:e6c613aa-6c8b-468c-b2eb-a8458578f94b/Oliver_Wyman_VDA_Automotive_supplier_report_final.pdf?mode=view.↩︎

8. Engelking, Niklas Jan. “Cariad: VW-Tochter stellt eigene Software-Entwicklung weitgehend ein.” heise online, October 4, 2025. https://www.heise.de/news/Cariad-VW-Tochter-stellt-eigene-Software-Entwicklung-weitgehend-ein-10711997.html.↩︎

9. Raasch, Philipp. “Der Aufstieg Und Fall (?) Von CARIAD – 5 Gründe Für Das Scheitern Der VW-Tochter.” LinkedIn, July 21, 2024. https://www.linkedin.com/pulse/der-aufstieg-und-fall-von-cariad-5-gr%C3%BCnde-f%C3%BCr-das-scheitern-raasch-ivgie/.↩︎

10. Dakić, Pavle, Igor Stupavský, and Vladimir Todorović. “The Effects of Global Market Changes on Automotive Manufacturing and Embedded Software.” Sustainability 16, no. 12 (2024). https://doi.org/10.3390/su16124926.↩︎

11. Shirokinskiy, Konstantin, and Markus Baum. “Driving the Future: Commercializing Automotive Software.” Roland Berger, September 19, 2024. https://www.rolandberger.com/en/Insights/Publications/Driving-the-future-Commercializing-automotive-software.html.↩︎

12. Schlosser, Joachim, and Jens Petersohn. “Maintaining Open-Source Based Software or What Is the True Cost of Free?” In ELIV 2023: Electric / Electronics for Commercial Vehicles 2023, edited by VDI Wissensforum, vol. 2423. VDI-Berichte. VDI Verlag, 2023. https://doi.org/10.51202/9783181024232-267.↩︎

13. Benjamin Fischer, Luis Karcher, Konstantin Kreye, and Florian Hacker. Dienstwagen auf Abwegen. Nos. 67-2021-DE. 2021. https://www.agora-verkehrswende.de/veroeffentlichungen/dienstwagen-auf-abwegen/.↩︎

14. Ruth Blanck. Klimaschutzinstrumente Im Verkehr. Dienstwagenbesteuerung. Umweltbundesamt, 2022. https://www.umweltbundesamt.de/sites/default/files/medien/366/dokumente/uba-kurzpapier_dienstwagenbesteuerung_kliv_v2.pdf.pdf.↩︎

15. Carl-Friedrich Elmer and Benjamin Fischer. “Dienstwagenprivileg: Mehr als ein Bauchgefühl.” Agora Verkehrswende, Oktober 2022. https://www.agora-verkehrswende.de/veroeffentlichungen/dienstwagenprivileg-mehr-als-ein-bauchgefuehl.↩︎

16. Peter Valdes-Dapena and Ivory Sherman. “Electric Car Timeline.” CNN, October 10, 2019. https://edition.cnn.com/interactive/2019/07/business/electric-car-timeline/index.html.↩︎

17. Diehl, Katja, and Doris Reich. Autokorrektur: Mobilität für eine lebenswerte Welt. FISCHER Taschenbuch, 2023.↩︎

18. IEA. “Global EV Outlook 2025.” May 14, 2025. https://www.iea.org/reports/global-ev-outlook-2025.↩︎

19. Schlosser, Joachim. “Always Be Curious – A Mindset for Better Conversations.” Dr. Joachim Schlosser, September 18, 2014. https://www.schlosser.info/mindset-conversation/.↩︎

20. Schlosser, Joachim. “Lesen über Sinn und Bestimmung: Simon Sinek – Start with Why.” Dr. Joachim Schlosser, July 2, 2015. https://www.schlosser.info/frag-erst-warum-start-with-why/.↩︎

21. Sinek, Simon. Start with Why: How Great Leaders Inspire Everyone to Take Action. Portfolio, 2009. https://amzn.to/2YkDJSa.↩︎

22. Mattausch, Alexander, Joachim Schlosser, and Moritz Neukirchner. “E/E Architectures and the Automotive OS.” In 23rd Stuttgart International Symposium – Automotive and Engine Technology Documentation, vol. 1. Springer Vieweg, 2023. https://link.springer.com/chapter/10.1007/978-3-658-42048-2_13.↩︎

23. Senge, Peter M. The Fifth Discipline: The Art and Practice of the Learning Organization. Rev. and Updated ed. A Currency Book. Currency Doubleday, 2006.↩︎