The automotive industry is facing fundamental change: digitalization, artificial intelligence (AI) and global developments are changing the rules of the game. In a recent podcast, hosts Patrick and Luca talk to automotive expert Thorsten - Head of Central Development at ZF and former Audi engineer - about these developments. The conversation sheds light on how digital technologies and AI are revolutionizing the automotive industry and which geopolitical challenges play a role in this.
Thorsten starts by briefly introducing himself: With 15 years of experience at Audi in areas such as driver assistance and vehicle safety and now nine years at ZF, he brings extensive expertise to the table. He was already experimenting with AI in the 1990s - long before the current wave of applications was possible. This expertise forms the starting point for an in-depth conversation about the future of mobility.
The podcast focuses on three major topics. Firstly, it deals with digitalization in the automotive industry: How is software changing vehicles and the development processes behind them? Secondly, AI takes center stage - from virtual sensors to intelligent driving functions. Thirdly, geopolitical influences will be discussed, such as how international partnerships or trade conflicts affect the industry. The moderators will begin by giving an overview of why these topics are so explosive right now. Current events - from new technologies such as ChatGPT to political decisions around emissions limits or trade restrictions - ensure that digitalization, AI and globalization are at the top of the automotive industry's agenda.
In this introduction, the course is set for the conversation: Everyone involved agrees that the automotive industry is in a phase of upheaval. Whether it's new business models, accelerated development cycles or competition with "new savages" from China - the changes are noticeable everywhere. The following summary provides a structured overview of the most important insights and discussion topics from the podcast.
Important insights
Software is becoming the key differentiator: The interviewees emphasize that cars are increasingly being defined by software and digital functions. In electric vehicles, the classic combustion engine is taking a back seat - instead, connectivity, user experience and digital features are deciding the brand essence of a vehicle. Manufacturers must therefore invest in software skills to raise their profile and achieve the industry's "iPhone moment". This means that cars of the future will be differentiated primarily by intelligent functions, regular updates and personalized experiences, rather than mechanical features.
Artificial intelligence unfolds enormous potential: Thorsten impressively describes how AI brings benefits both in vehicle development and in the end product. As early as the development process, AI models can help to carry out complex simulations or create a digital twin in order to develop components faster and more cost-efficiently. In the finished car, AI-based functions such as virtual sensors enable additional safety and comfort - for example, by combining data from several sensors to provide information that individual sensors alone could not collect. AI thus contributes to increasing quality and reducing costs and opens the door to innovations such as autonomous driving or predictive maintenance.
Geopolitical factors influence strategy and partnerships: The panel discusses how global developments are challenging the automotive industry. Trade conflicts and export restrictions - such as the US regulation limiting high-tech chips from NVIDIA for China - are forcing suppliers and manufacturers to re-evaluate their supply chains and partnerships. Thorsten reports on the need to play through alternative scenarios at an early stage: Which technologies and partners are available in the long term? At the same time, politicians are also reacting: the EU is promoting key topics such as autonomous driving and software platforms, while emission limits and data protection requirements vary from region to region. Companies need to adapt their strategy flexibly in order to reconcile both technological innovation and regulatory requirements worldwide.
Software Defined Vehicle as a model for the future: A key buzzword in the podcast is the "Software Defined Vehicle". The idea behind this is that the vehicle architecture of the future will be software-centric. Functions will be determined by code rather than hardware and can be subsequently installed or improved via updates. This development requires new ways of thinking: car manufacturers must decide which software components they develop themselves (in order to distinguish their brand) and where they rely on suppliers or open source solutions. The experts agree that no one can reinvent all software alone - cooperation and shared platforms are becoming increasingly important so that the industry does not fragment into countless isolated solutions.
Open cooperation and open source are becoming increasingly important: It becomes clear in the conversation that open source software is no longer a foreign concept for the automotive industry. Thorsten mentions initiatives in which suppliers (such as ZF) and partner companies work together on open platforms. Through open source, the industry can access the full innovation potential of the software world - which is necessary to keep pace with rapid development. At the same time, however, this also creates new competitors and issues of competitiveness: when core components are shared openly, companies have to add their value through special expertise, integration capability or services. New business models are emerging, for example through company-owned software subsidiaries or alliances that drive joint developments forward.
Accelerated development cycles are a must: One clear learning from the podcast is that traditional development times need to be drastically shortened. Where 36 months used to be planned for a new system, digitalization and international competition now demand more agile processes. The Chinese market in particular is setting standards: numerous manufacturers there are launching innovations at a rapid pace. German and European companies are feeling the pressure to establish faster product cycles and iterative development approaches (e.g. using simulation, digital twins and agile software development). If you are too slow, you risk losing touch with the market and changing customer expectations.
Customers expect a seamless digital experience: The new generation of car buyers values connectivity, updates and smart features rather than just horsepower figures. The podcast uses the example of a modern electric car: Young buyers are more likely to ask for app integration or driver assistance functions than cylinder numbers. Brands such as Volvo demonstrate this by installing an arsenal of sensors and computing power in the Polestar 2 in order to digitally realize their promise of maximum safety. The central question is: What will differentiate an electric car of the future for the customer? The answer increasingly lies in software functions and digital services that can be constantly expanded.
Traditional strengths need to be rethought: Despite all the digitalization, classic brand values remain relevant - they just translate into new technologies. Thorsten reports on how established car brands are now using software to emphasize their brand essence (such as safety at Volvo or "driving pleasure" at BMW). This means, for example, that a manufacturer with a safety image invests particularly in sensor technology, assistance systems and AI. The challenge is to transfer these established identities into the digital future. Success will come to those who combine traditional engineering skills with modern technologies and thus offer customers familiar quality in a new form.
The influence of global politics creates parallel worlds: One important realization is that there may no longer be "one" global car technology in the future. World political tensions mean that different standards could apply in China than in Europe or the USA - be it for operating systems, data protection or hardware. This increases the complexity for internationally active companies: software may have to be secured multiple times and adapted to regional requirements. Thorsten warns that this fragmentation of the technology landscape increases development effort and costs. Nevertheless, the European industry must find ways to maintain its innovative strength in this environment, for example through its own alliances and by doing its digital "homework" more quickly.
New players and cultural change in the industry: In conclusion, the interview shows that the automotive industry is becoming more open to external influences. Tech companies and start-ups are entering the field with fresh ideas - from smartphone manufacturers building cars to software start-ups with specialized AI solutions. For traditional companies, this also means a cultural change: collaboration on an equal footing, continuous learning and a certain humility that good ideas can also come from outside. Thorsten describes how ZF is questioning and improving its own processes through cooperation with Chinese "young and wild" companies and start-ups. This change ultimately makes the industry more dynamic and versatile, which benefits both consumers and the companies themselves if they embrace it.
Detailed key topics
The role of AI in automotive development and production
Artificial intelligence has long since arrived in the automotive industry and was highlighted in the podcast as a key factor for future success. Thorsten - who himself tried out neural networks back in the 90s - emphasizes how drastically the possibilities have expanded. Today, AI can bring enormous time and quality gains in vehicle development. One example is simulated ("virtual") sensors in the development process: instead of building expensive physical prototypes for each function, certain sensor values can be predicted using AI and simulation. This allows engineers to test systems before hardware even exists. AI thus creates a kind of digital laboratory in which countless variants can be run through - faster than would be possible in the real world.
AI is also finding its way into production itself. Whether in the production line, where algorithms carry out quality checks, or in the supply chain, which is optimally controlled using AI analysis - learning systems are ensuring efficiency everywhere. In the podcast, we discuss how AI shines particularly when classic mathematical models reach their limits. AI can master complex physical relationships that are almost impossible to calculate by recognizing patterns. The result is more intelligent production processes and more robust products. For example, predictive maintenance can prevent breakdowns in the factory because AI models detect anomalies at an early stage. In vehicle development, in turn, AI makes it possible to learn from large amounts of data - such as from vehicle sensors or tests - and constantly optimize the design.
In modern vehicles themselves, AI is increasingly becoming an invisible passenger. The panel discussion outlines how driving assistance systems and autonomous functions are heavily dependent on AI. Systems such as emergency brake assistants or lane departure warning systems use neural networks to evaluate camera images and other sensor data in fractions of a second. One exciting concept that Thorsten explains is the aforementioned virtual sensor: by combining various physical sensors (e.g. camera, radar, lidar), AI can create a kind of new "metasensor" that provides a more comprehensive picture of the environment than each individual sensor on its own. In theory, this could even eliminate the need for a physical sensor without sacrificing accuracy - for example, by using a camera to partially assume the function of a rangefinder with the help of AI. In practice, people are cautious here (the limits of physics cannot be completely overridden), but the principle shows where the journey is heading: AI expands the capabilities of vehicles and makes driving safer and more comfortable.
Another aspect is personalization through AI. Intelligent assistance systems that adapt to the driver are conceivable - even if it was jokingly noted in the podcast that not everyone wants to dictate a diary via ChatGPT in the car. For example, AI systems could learn the driving style and optimize safety systems or navigation accordingly. Overall, there was a consensus that AI will be one of the main drivers of innovation in the coming years, from the development office to the road.
Challenges of digitalization in the industry
Digitalization presents the automotive industry with a variety of challenges. In the podcast, it becomes clear that it is not just about new technology, but about a comprehensive change in processes and culture. One of the biggest hurdles: the long-established development cycles. Traditionally, vehicle or platform developments took several years - but this is too slow in the digital age. When software becomes the central component, customers expect annual improvements, similar to smartphones. Thorsten makes it clear that companies need to find ways to significantly accelerate their innovation cycles. Methods such as agile development, early simulation (keyword Digital Twin) and continuous updates after the sales launch will have to become standard in order for manufacturers to keep up.
Another challenge is the integration of new skills into a traditionally mechanical industry. Suddenly, software developers, data scientists and AI experts are in demand - profiles that used to be rare in car manufacturing companies. This raises questions: How do you attract these talents, and how do you change internal structures so that software takes on an equal role alongside engine and body construction? In the podcast, Thorsten explains that ZF recruits young talent directly from universities as well as bringing experienced specialists into the team. Collaborations with universities and an attractive innovation environment (such as projects in AI and high-performance computing) help to bring new skills into the company. Nevertheless, the digital transformation also requires a cultural change: away from silo thinking and towards interdisciplinary collaboration, in which a software update is treated just as importantly as a new gearbox.
Alongside this is the challenge of breaking down old ways of thinking. The presenters talk humorously about the "arrogance" of some manufacturers in wanting to do everything better themselves - alluding to companies that would prefer to develop every aspect internally in order to create their own "iPhone moment". However, the reality of digitalization shows that collaboration is often more efficient. No car manufacturer can single-handedly reinvent and perfectly maintain every software component. Instead, partnerships with tech companies (e.g. for cloud services or AI platforms) and with suppliers are necessary to cope with the enormous workload. This paradigm shift - from isolated tinkering to networked development - is not easy for many traditional companies, but it is unavoidable.
In addition, digitalization means that customer feedback and data are continuously incorporated into further development. Vehicles today generate vast amounts of telemetry data. The challenge is to use this data securely and in compliance with data protection regulations in order to improve the product. European manufacturers must also take into account the strict data protection requirements of their home markets (keyword GDPR), which makes the use of data challenging. All in all, digitalization demands unprecedented flexibility from the automotive industry - in terms of technology, personnel and organization. Those who manage to break down rigid processes and think in terms of software will lay the foundations for keeping up in the digital race.
The influence of geopolitical factors on the automotive industry
Globalization and politics were also a key topic of the podcast. The automotive industry operates globally - and is therefore susceptible to geopolitical tensions and regulations. Thorsten provides insights into how his company proactively deals with such influences. One striking example is the trade disputes between the USA and China: as a supplier that cooperates closely with chip manufacturer NVIDIA, for example, ZF witnessed how the US government imposed export restrictions on certain high-performance chips. Such decisions force companies to have contingency plans. What do you do if an important technology partner is suddenly no longer allowed to supply its products to certain markets? During the discussion, it became clear that supply chains must be globally diversified and robust at the same time in order to mitigate political risks.
Europe itself also sets political framework conditions that influence the industry. One current keyword was the debate about CO2 limits. While many suppliers and manufacturers are working flat out on low-emission technologies and sustainability, there has been talk at times that regulators could relax emissions standards in order to cushion energy crises, for example. In the podcast, the question arises: will this slow down innovation in green technologies? Thorsten argues that the industry should focus on long-term trends regardless of short-term political relief. Electromobility and CO2 reduction remain on the agenda - even if the electric revolution is slower than hoped, hybrid solutions and efficiency improvements are still in demand. This shows that politics can change the rules of the game, but successful companies will not be thrown off course by every twist and turn, but will strategically stay on course for future technologies.
Another geopolitical aspect is the competition of systems: China, the USA and Europe could go their own ways in some digital areas. In the podcast, the participants mention different operating systems for cars as a possible scenario. While Western manufacturers may rely on Linux variants or QNX, China is developing its own software stacks, and all of this has to work in different jurisdictions. This makes the development work complex. Thorsten describes that it is likely that different software platforms will have to be mastered in parallel - an effort that can only be managed with strong teams and resources. At the same time, he believes that Europe will have to focus on its strengths. The tradition of close cooperation between car manufacturers (OEMs) and large suppliers - a model that has produced many innovations in the past - could be a key to remaining competitive in uncertain times. As international cooperation becomes more difficult, regional alliances may come to the fore again.
Not least, China as a market and driver of innovation is a recurring theme. The hosts ask Thorsten what it's like to work with the "young guns" from China. His answer: it's different, but enriching. Chinese companies are extremely fast and agile, which encourages Western partners to speed up their own processes. The learning curve is two-way - you can learn just as much from the newcomers from the Far East as from the established OEMs in the West. However, politics in China play by their own rules: Access to the market, local regulations or requirements (e.g. for digital services or data protection) sometimes differ greatly. For global suppliers, this means operating in different worlds in parallel. Overall, the podcast participants conclude that although geopolitical influences present the automotive industry with difficult challenges, they can also drive innovation. They force you to always have a plan B in your pocket and to constantly adapt to new conditions - a challenge that has almost become the norm in times of rapid technological change.
Software Defined Vehicle - The future of vehicle technology
The term "Software Defined Vehicle" (SDV) comes up particularly often in the podcast - no wonder, as it is seen as a vision for the future generation of vehicles. The idea behind it is that a car is primarily defined by its software capabilities, while the hardware becomes a platform that can support different software as flexibly as possible. Thorsten and the presenters paint a picture of a vehicle that is constantly evolving over its life cycle as updates unlock or improve new functions.
A practical example: Where today the performance of a car is dominated by engine power or chassis, tomorrow digital features could take center stage. An electric car inherently has similar drive characteristics - what makes it different is the user interface, autonomous driving functions, networking with smart home or other services and the like. These functions are made possible by software and can theoretically also be retrofitted via over-the-air updates. The podcast discusses how manufacturers deal with this: Should they try to develop this software completely in-house to make their brand distinctive? Or is it smarter to rely on shared platforms and only differentiate themselves in certain areas?
Thorsten describes a balancing act here. Some manufacturers - BMW, for example - want to make a lot of software themselves in order to anchor their brand essence ("the joy of driving") in every digital pore. Others are more open to using standard software from suppliers so as not to reinvent every wheel. ZF itself, for example, has developed a product called Cubix, a type of universal control software for chassis components. A car manufacturer could buy this in and concentrate on writing the software parts themselves that really make up their unique selling point. Ultimately, the consensus is that the car of the future will be a mixture of both: there will be components that are similar across manufacturers (comparable to an operating system), and others that vary depending on the brand and shape the identity of the vehicle.
However, technical requirements are also necessary for a software-defined vehicle to become a reality. The electronic architecture of cars must change: Away from many scattered control units towards high-performance computers in the car that act like the "brain". In addition, there are powerful system-on-chip (SoC) components - the podcast mentions NVIDIA's Orin chips, for example, which are installed in current electric cars to process masses of data for assistance systems. Equally important is a seamless connection to the cloud. The concept of SDV assumes that some calculations or services take place outside the vehicle (in data centers), while critical functions remain local to the car (edge computing). This cloud-edge combination makes it possible to continuously improve computationally intensive AI models, for example, without having to update each car individually, or to evaluate large volumes of data centrally.
The podcast makes it clear that the industry is undergoing a radical change here. Consortia have been formed (Thorsten mentions one that is probably working on such architectures across industries) to develop common standards for the software-driven car. At the same time, every manufacturer wants to avoid becoming completely dependent on tech giants - the car should not become a mere appendage to the smartphone, but a digital product in its own right. The future of the software-defined vehicle therefore requires both cooperation and competition. All in all, the discussion partners are certain: in five to ten years, software will determine the success or failure of a car manufacturer. Those who can offer a stable, flexible software platform will be one step ahead.
Open source and new business models
Closely linked to the topic of software platforms is the question of open source and new business models. In the podcast, there is an exciting discussion about how much openness the automotive industry can and must allow. Thorsten mentions that his company is already taking steps in this direction: ZF has founded a subsidiary together with partners (including the software company KPIT) that deals specifically with open source software for vehicles. The idea behind this: To jointly develop certain basic technologies in order to avoid duplication of work and create a quasi-standard that everyone can build on.
Open source in the automotive industry means that companies make some of their code public and release it for shared use. The advantage is obvious: the innovative power of the global developer community can be harnessed. Instead of each manufacturer writing their own operating system or AI library, for example, resources could be pooled. But there are also obstacles: finding business models in an open ecosystem is challenging. After all, every company wants to earn money. The podcast sheds light on this dilemma. One solution could be for the basic platform to be open, but for specialized additional functions or services to generate income. Similar to Android in the smartphone sector: the operating system is open source, but device manufacturers earn money from the hardware and their own software extensions.
Thorsten and the moderators are optimistic that open source can lead to success in certain areas. Examples from the past, when the automotive industry created common standards (think GENIVI, AUTOSAR or other consortia), show that rivals can cooperate on fundamental topics to make life easier for each other. However, it is important to make clear agreements - also with regard to liability and security. If everyone is working on an open software basis, bugs must be fixed quickly and security gaps closed, as otherwise everyone would be affected. This requires new forms of collaboration, such as foundations or alliances in which competing companies pull together.
New business models are emerging in this context, particularly for suppliers. A supplier could change from a pure hardware supplier to a software service provider that supports an open platform and sells support and integration services for it. Or manufacturers launch their own software products on the market that can also be used by competitors - an unusual idea in an industry that has mainly traded physical parts for decades. The podcast gives the reader the feeling that a lot is currently in motion here. Openness should no longer be a taboo if it is the key to rapid innovation. At the same time, every company must define where its distinctive features lie that it does not want to reveal. The balance between open sharing and closed protection will be exciting to observe in the coming years.
Change in the automotive industry and future trends
Finally, the podcast discussion turns to the big picture: What does the automotive industry of tomorrow look like, what trends are emerging? One key point is the cultural change I mentioned earlier. Tradition and innovation coming together - what initially sounds like a conflict will, at best, lead to a new self-image for the industry. Thorsten talks about the need to constantly question one's own efficiency and working methods, driven by competition and new impulses. This continuous improvement process has always been part of industrial culture, but in the age of software and globalization it takes on a new urgency. Companies need to become more adaptive, experiment and sometimes risk cutting off old habits.
A trend that is clearly recognizable: Electromobility will come, but perhaps in a different way than originally thought. Expectations that combustion engines would disappear completely within a few years have been put into perspective. Instead, experts see a transition phase with hybrid technologies and a diversification of drive systems. This is also necessary in order to do justice to different markets and infrastructure conditions worldwide. But in the long term, there is hardly any way around electrification, especially as it goes hand in hand with digitalization. An electric car is ideally suited to receiving new functions via software, and it is easier to integrate with electronic control systems than a complex combustion engine. The future trends of electrification and digitalization are therefore mutually dependent and will shape vehicles together.
Autonomous driving also remains on the agenda, albeit with a more sober outlook. The podcast admits that fully autonomous driving (Level 5) still raises many unanswered questions - whether in terms of technology or liability. Nevertheless, partial automation and assistance systems have made enormous progress. There is a trend towards more integral safety: car manufacturers such as Volvo are packing so many sensors and computing power into the vehicle that they act as a kind of safety net, even when a human is at the wheel. The vision is accident-free driving in which computers can iron out human errors. Step by step, technology is getting closer to this goal, even if the fully autonomous robotaxi future still seems a little way off.
An often underestimated trend is the merging of the technology sectors. Cars are becoming part of a larger digital ecosystem. For example, the podcast participants discuss the fact that future buyers will choose their vehicle based on digital criteria: Does the car fit seamlessly with my smartphone? Does it offer streaming, smart home connectivity, regular feature updates? Manufacturers are therefore increasingly working together with IT companies (think of collaborations such as BMW with Microsoft or Volkswagen with Google for cloud and AI). This is giving rise to new car-related services: from subscription models for certain driving functions to mobility apps that connect different modes of transportation. The traditional boundary - here the car, there the rest of the digital world - is becoming blurred.
Another aspect of the future is competition from new entrants from outside the industry. This has already been mentioned: Companies such as Xiaomi (known from electronics) or even Apple are repeatedly traded as potential car manufacturers. These companies bring a fresh perspective and immense financial strength. The traditional car industry is watching this with concern on the one hand, but with motivation on the other - because it forces them to remain innovative. Ultimately, the future could be characterized by collaborations where old and new come together: for example, a traditional OEM supplying production expertise and a tech group contributing software and electronics. Such alliances would have been unthinkable a few years ago, but are now quite realistic.
In summary, a look at the coming trends shows: Flexibility, openness and customer focus will determine success. The industry is changing from the ground up, but this change holds great opportunities. Companies that are prepared to reinvent themselves can play an active role in shaping the mobility of the future.
Conclusion
The podcast impressively illustrates the diversity of topics that are currently affecting the automotive industry. From the digitalization of development processes to AI in the end product - there are opportunities everywhere to increase efficiency and build better cars. At the same time, companies must learn to deal with uncertainties: Geopolitical tensions, new competitors and changing customer demands require a high degree of adaptability. A key insight from the discussion is that digitalization and AI are no longer optional extras, but will become a survival factor. Those who use these tools skilfully can not only reduce costs, but also differentiate their products and open up new business areas.
The importance of digitalization and AI for the future of the automotive industry can hardly be overestimated. Software and artificial intelligence will make vehicles safer, more environmentally friendly and more comfortable. For consumers, this means that cars will become "smart companions" more than ever before - they will assist with driving, adapt to personal preferences and always stay up to date thanks to updates. Digital technologies can also help in terms of sustainability, for example through more efficient driving or smarter traffic management that reduces congestion. For consumers, the change therefore brings tangible benefits, from greater safety to new services, but must also be transparent and trustworthy (keyword: data protection and reliability of systems).
For companies, the digital transformation also opens up new horizons, but requires courage and investment. Manufacturers and suppliers can tap into additional sources of income with the new technologies - for example through software subscriptions, data-based services or partnerships with tech companies. At the same time, competitive pressure is high: those who miss the boat risk losing market share. The podcast makes it clear that the German and European players have nothing to hide: With their technical expertise and quality standards, they have a good chance of remaining leaders in the future. But only if they set the right course now - in other words, if they invest in innovations, promote talented employees and remain open to cooperation.
In conclusion, it can be said that the topics discussed in the podcast - digitalization in the automotive industry, AI and geopolitical challenges - are not separate from each other, but interlock like gears. Together, they are driving the transformation of mobility. For us consumers, this means exciting times: Tomorrow's car will be different from yesterday's. And for companies, it means laying the foundations today to be successful tomorrow. The podcast provides plenty of food for thought and shows that despite all the disruption, one thing remains the same: Change is the only constant, and those who actively shape it will ultimately benefit from it.