Live Blog from the AUTOMOBIL-ELEKTRONIK Kongress 2026

On 16 June 2026, the 30th AUTOMOBIL-ELEKTRONIK Kongress opens its doors in Ludwigsburg. Over two days, OEMs, suppliers, semiconductor manufacturers, software companies and technology providers will discuss the future of the vehicle. The agenda focuses on software-defined vehicles, AI in the vehicle, next-generation E/E architectures, semiconductor strategies, lidar, edge AI, open source technologies and the question of how technological concepts can be translated into scalable series-production solutions.

The anniversary theme connects 30 years of AEK with the major challenges and opportunities facing the industry. What role will AI play in future vehicle architectures? How will SDV platforms reshape collaboration between OEMs, suppliers and software partners? Which technologies can make the transition from demonstration, datasheet and pitch deck into real vehicles? And how quickly can the industry bring new functions safely, economically and effectively to the road?

Our live blog will accompany AEK 2026 with impressions from Ludwigsburg, insights from the stage, analysis of the key topics and highlights from across the network. We will also be covering the event on LinkedIn.

Alexandre Corjon brings Edge AI into vehicle architectures

Dr Alexandre Corjon, Senior Vice President Engineering at Sonatus, closed the congress by highlighting the growing challenge of integrating Edge AI into modern E/E architectures.

For Corjon, it is not enough to simply add AI as another feature on top of existing systems. Intelligence must be available within the vehicle itself, as latency, bandwidth constraints, privacy requirements and response times make it impossible to rely solely on cloud-based processing for many applications.

To illustrate this, he described three compute layers within the vehicle: sensors and intelligent endpoints, zonal gateways and central high-performance computers. Depending on the application, AI can take on different roles across these layers, ranging from local data preprocessing and sensor fusion to event detection and the execution of larger models on central compute platforms.

A key theme of his presentation was the need to clearly separate AI workloads from safety-critical functions. The two have fundamentally different requirements in terms of memory, execution, determinism and fault tolerance, making architectural separation essential.

Beyond the architecture itself, Corjon argued that the industry needs platforms capable of collecting data efficiently, feeding insights back into cloud-based development environments, orchestrating edge performance and managing software and model updates throughout the vehicle lifecycle.

Among the examples he cited were in-vehicle fault analysis and user-centric personalisation, where the vehicle identifies behavioural patterns and proactively recommends suitable settings.

Immediately afterwards, Alfred Vollmer delivered the closing remarks, bringing a successful 30th AUTOMOBIL-ELEKTRONIK Kongress to an end. The event will return to Ludwigsburg near Stuttgart on 22 and 23 June 2027 for its 31st edition.

Liu Qiang describes the shift from smart vehicles to embodied AI

Dr Liu Qiang, Vice President of Li Auto and General Manager of the Li Auto Germany R&D Center, explored a question that extends beyond the software-defined vehicle: what comes next?

For Li Auto, the answer is embodied AI – systems that can perceive their surroundings, draw conclusions from them and act in the physical world. The company increasingly views the vehicle as an intelligent agent whose architecture mirrors perception, cognition, a nervous system and movement.

Liu outlined four core layers underpinning this vision. Sensors and 3D vision serve as the vehicle’s “eyes”. Large AI models provide perception, spatial understanding and decision-making capabilities as the “brain”. Proprietary inference chips deliver the required computing foundation, while chassis systems, brakes, steering and actuators form the physical interface to the real world.

Examples ranged from natural multimodal interaction through voice and gestures to scenarios in which a vehicle autonomously drives to a charging station and subsequently finds a parking space on its own.

Jørgen Behrens turns navigation into an AI companion

Jørgen Behrens, Vice President and General Manager of Google Maps Automotive at Google, placed the driver and cognitive workload at the centre of his presentation.

In his view, the quality of modern vehicles is increasingly reflected in the digital experience they provide – helping drivers stay oriented, reducing stress and making everyday journeys easier.

Behrens described Google Maps as evolving far beyond traditional navigation. The journey extends from mapping APIs and smartphone projection through Android Automotive and Google Built-in to Gemini-powered experiences and immersive navigation.

Three examples illustrated this direction. The first was a more realistic 3D navigation experience featuring detailed roads, buildings, landmarks and live lane guidance, helping drivers understand their surroundings and make safer decisions more quickly.

The second focused on electric vehicles. Through Energy Intelligence, Google aims to improve charging planning by incorporating consumption forecasts, battery state, charging locations and increasingly accurate predictions about charger availability upon arrival.

The third example centred on Gemini. Rather than relying on predefined voice commands, drivers will be able to ask more natural questions about routes, their vehicle, nearby locations or charging plans and continue the interaction through conversational dialogue.

Chengyin Yuan outlines China’s path towards semiconductor sovereignty

Dr Chengyin Yuan, Secretary-General of the China Automotive Chip Industry Innovation Strategic Alliance, provided insights into the evolution of China’s automotive and semiconductor markets. 

He showed how new energy vehicles, autonomous driving functions, data-driven development and centralised E/E architectures are significantly changing the requirements for automotive semiconductors. The focus is shifting towards greater compute performance, memory capacity, connectivity, functional safety, reliability, energy efficiency and new architectural approaches such as RISC-V and chiplets.

At the same time, Yuan made it clear that although China has made substantial progress in automotive semiconductors, the country still sees dependencies and risks in areas such as EDA tools, core IP, materials, manufacturing equipment and advanced semiconductor production.

Semiconductor sovereignty was therefore presented as a strategic prerequisite for future intelligent, connected and electrified vehicles.

Looking ahead, Yuan outlined a roadmap centred on higher compute performance, near-Level-5 autonomous driving scenarios, AI-powered cockpits, next-generation power semiconductors and more open models of industrial collaboration.

The presentation offered a distinctly Chinese perspective on the close relationship between vehicle architecture, AI, semiconductor strategy and industrial value creation.

Christian Sobottka turns the car into an experience platform

Christian Sobottka, President & CEO of Harman and President of Harman Automotive, shifted the focus away from individual technologies towards the outcome that ultimately matters: a differentiated, emotional and continuously improving user experience.

Achieving this requires a new level of orchestration. Cockpit systems, audio, displays, connectivity, ADAS, AI and contextual data need to be combined into a coherent and trustworthy overall experience.

As a result, the benchmark is moving away from individual features towards intelligent interactions that select the right modality at the right moment – whether through voice, haptics, visual interfaces or audio.

Sobottka also highlighted how strongly China and consumer technology ecosystems are reshaping expectations around speed, integration and digital experiences.

Harman’s response is built around productisation, reusable software and hardware building blocks, central compute platforms and AI-supported workflows. The goal is to bring new solutions to start of production significantly faster than before.

One of the most memorable points of his presentation was his argument against fear as a driver of change. Instead, organisations need to evolve their structures, collaboration models and decision-making processes in ways that foster passion, focus, courage and execution speed.

The result was a broader call to view the vehicle not as an isolated product, but as part of a larger digital ecosystem centred on the customer experience.

Alwin Bakkenes explains Volvo’s path towards a software-defined future

Alwin Bakkenes, Head of Software Engineering R&D at Volvo Cars, described the company’s transformation through three major challenges: electrification, the end of globalisation and growing hypercompetition.

For Volvo, this transformation is closely tied to its core brand value of safety. Real accident data, sensor data and millions of data points are intended to help create vehicles that recognise risks earlier, improve over time and receive new functions more quickly.

To achieve this, Volvo has fundamentally changed its vehicle development approach. The company is moving away from traditional domain architectures and rigid vehicle programmes towards a zonal architecture, central computers, connected vehicles and its own software factory.

A particularly important capability is the ability to build software continuously, test it virtually and roll it out across the entire vehicle fleet in fixed cycles. At the same time, the approach remains strongly industrial. Every new function must remain manageable across model years, ongoing vehicle programmes, test environments, plants and vehicles that have already been delivered. 

The presentation also offered an interesting look at partnerships, including with Google, and at the question of which parts of software should be developed openly, jointly or as proprietary technology.

Ahmad Bahai redistributes AI inside the vehicle

Dr Ahmad Bahai, Senior Vice President and CTO of Texas Instruments, continued the discussion on future vehicle architectures at semiconductor and system level. His starting point was that AI is changing not only software and central computers, but also sensors, power supply, data paths, timing and the question of where intelligence should actually reside in the vehicle. 

A particularly important principle for him was distributed intelligence. Not every piece of information needs to be sent in raw form to a central computer or the cloud. Instead, vehicles will require local processing directly at the sensor, edge AI and federated learning approaches.

Using radar and lidar as examples, he showed how strongly data volumes, synchronisation, bandwidth and energy efficiency influence architectural decisions. Bahai also sees a major need to scale AI accelerators, from tiny NPUs at the outer edge of the system to more powerful edge platforms.

The resulting picture of the vehicle is less that of a single supercomputer on wheels and more that of a distributed intelligent system. Central computers, sensors, local AI building blocks, networks, power management and cloud connectivity must be brought into the right balance depending on the application.

Peter Schiefer shows why architecture will shape the future of mobility

Peter Schiefer, President and CEO of the Automotive Division at Infineon Technologies, brought the second day of the congress back to the hardware side of the software-defined vehicle. 

For him, the future does not simply lead to one large central computer handling every task. Instead, vehicles will need a deliberate distribution of compute power, safety, energy efficiency and local intelligence – from central compute units to smart sensors and actuators.

Ethernet, intelligent endpoints, microcontrollers, power management and secure communication paths are therefore becoming central building blocks of the SDV. Schiefer placed particular emphasis on the role of microcontrollers. In his view, they remain the safety backbone of the vehicle, continue to control many local functions and must at the same time become more powerful, more connected and better prepared for edge AI applications.

Infineon also sees RISC-V as an important step, as open standards and scalable ecosystems become more relevant for future vehicle platforms. Europe, Schiefer argued, still has strong opportunities through systems expertise, semiconductor know-how, investment in manufacturing and close cooperation between the automotive and semiconductor industries.

Markus Heyn rethinks the vehicle of the future

Dr Markus Heyn, Member of the Board of Management and Chairman of the Mobility business sector at Bosch, broadened the discussion on the second day from the software-defined vehicle to the question of how future vehicles can still be developed at scale.

The pressure is clear: China is setting new benchmarks in innovation speed, user experience and cost, while global markets are also drifting further apart. For Bosch, this results in three priorities: better user experience, greater speed and significantly lower cost.

This will require seamlessly integrated functions across domains, open software ecosystems and new forms of collaboration, including around Eclipse S-CORE and SDV platforms.

At the same time, Heyn made clear that hardware continues to play a central role. Centralised compute architectures, modular chiplet concepts, by-wire systems, 48 V architectures and new energy management approaches are intended to help reduce complexity and cost. One of the biggest challenges, in his view, remains global scaling, because regional regulations, data rules, software stacks and supply chains are increasingly forcing different technology paths.

His conclusion was that anyone who wants to develop future vehicles that are affordable, innovative and close to customer needs must think about software, hardware, energy architecture, partner networks and local market requirements together.

Audi China opens day two with a look at China’s innovation momentum

The second day of the AUTOMOBIL-ELEKTRONIK Kongress opened with a presentation on the electrification and digitalisation of China’s automotive market. At the centre was the question of why China has long since become more than just a major sales market for the industry.

Local customer expectations, government frameworks, strong tech ecosystems, high development speed and an enormous talent pool are now setting the pace of innovation there.

Ivo Muth, Executive Vice President of Audi China R&D, Audi (China) Enterprise Management, made clear how strongly business models and development processes need to change when global architectures meet local requirements, Chinese speed and new digital ecosystems.

He broadened the perspective beyond the car. Physical AI, intelligent vehicle functions and humanoid robots could trigger further disruption in the future, using technologies that are in some cases very close to those found in automated vehicles.

A day between SDVs, AI and 30 years of AEK

The first day of the 30th AUTOMOBIL-ELEKTRONIK Kongress in Ludwigsburg showed how strongly the industry is reorganising itself around software-defined vehicles, AI, semiconductors, data and new organisational models. Ricky Hudi opened the day by looking back on 30 years of AEK and handing over his role as Chairman of the Advisory Board to Alfred Vollmer.

A clear thread then ran through the day: the automotive industry needs to bring software, electronics, semiconductors, data and organisation much closer together if concepts are to become series-production solutions. Again and again, the discussion returned to how complexity can be managed – in vehicle architectures, supply chains, software ecosystems, wiring systems, sensor technology and industrial AI.

It also became clear that technological capability alone is not enough. Speed, collaboration, standardisation, customer proximity and the ability to learn from data more quickly will become decisive.

The AUTOMOBIL-ELEKTRONIK Kongress continues tomorrow. We wish all readers a pleasant evening and look forward to welcoming you back to the live blog tomorrow.

Joachim Langenwalter calls for the shift to the AI-defined vehicle

Joachim Langenwalter, CEO and Founder of TMT Co-Pilots, closed the conference day with a clear message: in his view, the industry is leaving the era of the software-defined vehicle and entering the age of the AI-defined vehicle.

For him, data has become the decisive lever. AI projects, he argued, often fail not because of technology limitations, but because of poor data quality, organisational shortcomings and slow decision-making processes.

China featured prominently in his assessment. According to Langenwalter, the focus there has already shifted towards customer-defined vehicles, rapid learning cycles, digital ecosystems and direct feedback loops from the market. European companies still benefit from strong brands, engineering expertise and vast data resources, but they need to translate those strengths into new functions and better customer experiences much faster.

That requires fewer organisational silos, more cross-functional teams, stronger supplier integration and a culture that values learning and adaptation more highly than long-term forecasting. His appeal to the industry was straightforward: anyone who wants to shape the future of the car must think about organisation, data and AI as a single challenge – and act now.

Dominik Wee: When code becomes a commodity

Dominik Wee, Corporate Vice President of Forward-Deployed Engineering at Microsoft, brought the software discussion to a new level in the late afternoon. His central question was: what happens when AI accelerates software development to the point where writing code is no longer the primary bottleneck?

For the automotive industry, he highlighted a particularly interesting implication. The sector’s traditional strength in specifications, standards, requirements management and V-model development processes could become a major advantage in the AI era. AI agents perform better when they operate within clearly defined guardrails, context and requirements.

Wee therefore spoke about the rise of “spec-driven development”. In this model, specifications become more important, while code increasingly becomes the output of a well-managed engineering process rather than the primary focus.

He also argued that the V-model itself is evolving, as AI agents become capable of linking requirements, code bases, compliance checks and remediation proposals much more closely.

Despite these changes, domain expertise remains essential. Engineers who understand the system, the customer requirements and the operational context will be better positioned to guide AI agents effectively.

Wee compared the current transformation to the transition away from hand-written assembly language. Programming as it is known today, he suggested, could lose its central role far more quickly than many expect.

Glen de Vos explains why Lidar 2.0 has to arrive now

After looking at the “brain” of vehicle occupants, the “Technologies to Watch” format turned to the eyes of the vehicle: Glen de Vos, CEO of MicroVision, spoke about “Lidar 2.0”.

His starting point was a sober look back at Lidar 1.0. Since the early 2010s, the technology has been regarded as an important enabler for automated driving and robotaxis, but it has so far only partially met the high expectations placed on it in the mass market.

For de Vos, the reason lies less in the basic principle of the technology than in cost, industrialisation and the absence of a compelling customer benefit. Technologically, he pointed to solid-state approaches, software-centred processing and stronger scaling through chip and wafer processes. 

He also emphasised the importance of multimodal perception: lidar, radar and cameras should work together to create a more robust view of the surroundings and provide mutual redundancy. For Level 3 functions and beyond in particular, lidar could become more relevant again if the sensors become cheaper, smaller, easier to integrate and better aligned with vehicle architectures.

Niall Berkery looks inside the brains of vehicle occupants

With “Technologies to Watch”, AEK moved into a more compact format in the late afternoon: two technologies, two short impulses and no Q&A session on stage.

The first presentation came from Niall Berkery, CEO and Co-Founder of Neumo, with a topic that initially sounds futuristic: brain-wave sensing in the vehicle. 

“The real story of a driver’s state lies in the brain,” Berkery said. Neumo wants to use EEG signals for this, but without a cap or wearable device. The sensor is positioned in the headrest area and is designed to detect brain activity from a short distance.

AI is then used to turn these signals into usable indicators of fatigue, cognitive load, relaxation or attention. This is particularly relevant for driver state monitoring and automated driving functions, for example when a vehicle needs to assess whether the human driver is ready to take back control.

Berkery also sees potential for wellness and personalisation functions in the cabin – with a car that better understands the state of its occupants and can respond accordingly.

The second “Technologies to Watch” impulse will focus on “Lidar 2.0”, presented by Glen de Vos, CEO of MicroVision.

Julian Raabe and Jörg Tischler show the value of industrial AI

Dr Julian Raabe, Senior Vice President and Head of Automotive & Manufacturing at T-Systems, and Jörg Tischler, Vice President Connected Mobility at T-Systems, picked up several of the day’s central themes in the afternoon: AI, global disruption, growing data volumes, software scaling and secure, sovereign infrastructure.

According to them, the decisive factor is connecting development, production, quality, vehicle and customer data. This requires a robust infrastructure that provides access to models, supports processes and at the same time addresses security, governance and cost.

Examples included simulations, digital factory planning, synthetic data and automated vehicle movements in plant and logistics environments. One point became particularly clear: AI should not only accelerate individual work steps, but make entire value chains more integrated, more resilient and easier to control. Achieving this requires a combination of edge computing, cloud, connectivity, secure data spaces and AI platforms suitable for industrial use.

Walter Glück: E/E architectures are also defined by manufacturing

Rather than focusing solely on technical capabilities, Walter Glück of Leoni explored how future E/E architectures can be designed to become easier and more efficient to manufacture. At the centre of his argument was the wiring harness – a component that often remains in the background but has a major impact on vehicle cost, complexity, logistics and production. 

Today’s wiring harnesses remain highly complex, variant-rich and labour-intensive. In some cases, installation alone can take several hours. To address this, Leoni is working on more modular concepts, smaller functional units and highly automated production cells that could eventually achieve automation levels of up to 90 per cent.

According to Glück, reaching that goal will require more standardised hardware modules, new connectivity concepts and closer alignment between OEM manufacturing, wiring-harness architecture, logistics and software development.

Collaboration was another recurring theme. Glück stressed that the industry needs to come together earlier, share knowledge more openly and make decisions faster. His conclusion was that the E/E architecture of the future is not defined by software alone. It must also be manufacturable, scalable and optimised across the entire supply chain.

Aish Dubey: Chiplets scale time, not just transistors

Dubey deliberately began with a broader point than semiconductors alone: 1.3 million road deaths per year show why architectural decisions in the vehicle are not only technically relevant, but also socially significant.

For him, AI, safety, memory, data movement and product lifecycles must be considered together in future vehicle architectures. This is especially important because vehicles remain in use for many years, while AI models and software workloads change much more quickly.

Renesas therefore sees chiplets as a fundamental architectural decision. Different functions such as safety islands, AI compute, memory and I/O should be partitioned in a way that fits the respective technical, economic and safety-related risk. One central point was data movement. In future, the transistor alone will no longer be the main bottleneck. What matters more is how efficiently data flows and memory accesses are organised.

Dubey argued that chiplets can take the industry beyond traditional transistor scaling towards systems that can be qualified more quickly, adapted to new functions faster and used meaningfully for longer. At the same time, he made clear that this will require standards, collaboration and resilient ecosystems. His core message was that the industry must not allow itself to be driven only by software and AI. It also needs to think about silicon architecture early enough as the foundation for this transformation.

Thomas Irawan and Christian Salzmann put S-CORE into context

Dr Thomas Irawan of ETAS and Dr Christian Salzmann of BMW provided an update on the Eclipse S-CORE project, which was first introduced at AEK 2025. Their presentation focused on how open-source software can be developed for automotive applications in a way that delivers innovation while still meeting the industry's requirements for safety, security and series production.

One message stood out clearly: as software-defined vehicles become more complex, shared foundations across the industry are becoming increasingly important. S-CORE aims to address this challenge by creating reusable software building blocks and common standards that can be developed jointly by OEMs, suppliers and tool providers.

The presentation highlighted that automotive open source can only deliver its full potential when technical collaboration, governance structures and clearly defined responsibilities evolve together. In that sense, the session fit seamlessly into the broader themes of the congress: reducing complexity and creating a more robust foundation for software development in future vehicles.

Chiplets in the car: between modular design, business case and software complexity

Moderated by Dr Matthias Pillin, representatives from BMW, Stellantis, TSMC Europe, Bosch and SiMa.ai discussed where the technology already shows promise today – and why bringing it into vehicles is considerably more challenging than in other industries.

The appeal is obvious. Chiplets could help shorten development cycles, enable the reuse of individual functional blocks and accelerate the integration of new AI and compute capabilities into vehicle architectures.

At the same time, the discussion made clear that the automotive industry faces very different constraints from consumer electronics or data centre markets. Long product lifecycles, safety requirements, security considerations, validation efforts, cost pressures and volume economics all add significant complexity.

One theme ran through the entire discussion: software. As vehicle architectures become more centralised, complexity does not disappear – it shifts into software, integration and system validation. For OEMs, this means future hardware architectures increasingly need to follow software requirements rather than the other way around.

The panellists identified standardisation, clearly defined interfaces, robust ecosystems and capable system integrators as the key prerequisites for success. Their overall conclusion was that chiplets could become an important building block of future vehicle computing platforms – but only if technology, software, business models and production volumes align.

Ned Curic questions the term software-defined vehicle

Ned Curic, Chief Engineering and Technology Officer at Stellantis, opened with a provocative statement: he said he did not really know what a software-defined vehicle was supposed to be. After all, no one speaks of a “software-defined phone” or a “software-defined television”, Curic argued in essence.

For him, one thing remains decisive: it is still a car, albeit one with a great deal of electronics, a great deal of software and the opportunity to become more intelligent and more personal.

That was the focus of his presentation: how can the complexity of a group with 14 brands, numerous platforms, legacy architectures and grown software landscapes be reduced in a way that results in better vehicles?

Stellantis is addressing this through a more centralised architecture with STLA Brain, fewer control units, more computing power and significantly faster development cycles. Curic was particularly clear on software organisation: the traditional separation between hardware and software teams does not work everywhere. What matters are clear responsibilities, smaller teams, automation and closer collaboration with selected partners.

As an example, he cited Smart Cockpit, where Stellantis says it can reuse around 85 per cent of the software across different applications. His central point was that customers are not interested in terms such as SDV. Customers expect the car to work, feel personal, look good, be safe and feel right in everyday use.

Nakul Duggal: The car is becoming an AI system on wheels

Nakul Duggal, Group General Manager Automotive, Industrial and Embedded IoT at Qualcomm Technologies, focused his presentation on a question that is increasingly shaping the industry: how quickly AI is moving into the vehicle.

Right at the start, Duggal highlighted how rapidly the discussion has evolved. Just two years ago, few would have expected the industry to be talking so concretely about AI in the car. In his view, this is no longer a distant vision. The transition is already under way, driven by more powerful SoCs, new software stacks and AI models capable of running directly on the vehicle.

AI functions are no longer defined solely through traditional programming, but increasingly through models, prompts, data access and secure architectural frameworks. “If you use these terms in the context of the car, then you are having a conversation inside the vehicle, the vehicle sees the real world and responds to you,” Duggal said.

At the same time, he stressed that the key question is not only what is technically possible, but whether these systems can operate safely, reliably and transparently.

His conclusion was clear: AI in the vehicle is not something that will arrive in five years’ time. It is arriving now, and its full potential will emerge where compute platforms, software architectures, sensors and customer experience are developed as an integrated system.

Florian Weig explains why BMW has to rethink its supply chain

Dr Florian Weig, Senior Vice President Purchasing & Supplier Network Digital at the BMW Group, focused on the future of the automotive value chain at the 30th AUTOMOBIL-ELEKTRONIK Kongress. His starting point was a supply chain that delivers 36 million parts to BMW plants every day, but which now faces very different dynamics, risks and dependencies in the digital and electronics domain.

Weig showed how strongly the framework conditions have shifted over the past decade. Geopolitics, regulation, tariffs, export controls, technological restrictions and the competition for compute are changing OEM purchasing and partnership strategies.

He expressed the software shift particularly sharply: “We see that C++, the world’s leading programming language, is being replaced by the simplest language: English.” What he meant was the influence of AI on software development and value creation. Code alone is losing some of its differentiating power, while architecture, documentation, robustness, standardisation and ecosystem management are becoming more important.

According to Weig, three things will remain permanently decisive for BMW: innovation, better overall systems and low cost. As key levers, he cited standardised systems, vertical transparency and collaboration, as well as resilient designs that can cope with uncertain geopolitical conditions.

His conclusion was that Europe’s industry can only compete with strongly vertically integrated rivals if it consistently combines standards, open-source approaches, data ecosystems such as Catena-X and resilient partnerships.

Lars Reger makes the software-defined vehicle tangible

Lars Reger, Executive Vice President and CTO of NXP Semiconductors, tackled a topic that has accompanied the automotive industry for years: the software-defined vehicle. Right from the start, he addressed the fatigue that sometimes surrounds the SDV acronym – in essence, everyone talks about it, but the practical implementation often remains vague.

Rather than relying on buzzwords, Reger focused on architecture, building blocks and the challenge of bringing together high-performance processors, zonal architectures, safety, security and real-time vehicle data. A key element of his presentation was NXP’s S32N7 Super-Integration Processor, which he positioned as a foundation for what he called the “Artificial Intelligence Defined Vehicle”.

On the technical side, Reger emphasised that AI cannot be viewed in isolation. It requires a surrounding architecture, robust safety mechanisms and secure data pathways. His central message was clear: the vehicle of the future is not simply “AI on wheels”. Whether AI, data and compute can be transformed into a safe and scalable vehicle concept ultimately depends on the underlying system architecture.

Magnus Östberg: AI, collaboration and orchestration will shape the car of the future

Östberg's starting point was the combination of 140 years of automotive history, 30 years of the AUTOMOBIL-ELEKTRONIK Kongress and an industry that has repeatedly had to reinvent itself.

For Östberg, innovation must ultimately reach the customer as a product that creates emotion, fits the brand and delivers real value. A central theme of his presentation was collaboration. Using Bluetooth as an example, he described how a shared standard was able to become a technology with enormous market momentum. “If you remember just one thing, remember collaboration,” Östberg said in essence.

For Mercedes-Benz, the software-defined vehicle today means a combination of in-house software, open source, licensed software and partner integration. Östberg described AI as an existing force in the vehicle, from voice functions to automated driving. At the same time, he warned of the “jagged frontier”, the boundary between impressive AI capabilities and the continued need for validation.

His conclusion was that, alongside collaboration, orchestration, automation and compliance will become decisive in determining which functions run in the vehicle, in the cloud or through an interaction between both worlds.

Ricky Hudi looks back on 30 years of AEK and hands over to Alfred Vollmer

Ricky Hudi opened the 30th AUTOMOBIL-ELEKTRONIK Kongress with a personal reflection on the event’s three-decade history. The long-serving Conference Chair traced the evolution of automotive electronics from early, often fragmented stand-alone solutions through connected vehicles and the connected car era to software-defined vehicles and AI-driven architectures. In doing so, he highlighted how closely the congress has been linked to the industry's technological transformation.

Hudi recalled that the idea for AEK emerged in the 1990s as increasing vehicle connectivity created a growing need for a platform where the industry could exchange ideas and experiences. He also reflected on the industry's past relationship with semiconductor suppliers: “There was a dinner, there was a meeting, there was a handshake. And there was a commitment. That was it.”

The opening speech also carried a personal note. Hudi looked back on a career that has largely run in parallel with the development of AEK itself. After several years as Chairman of the Advisory Board, he announced that he would step down from the role. “After careful consideration, I have decided to step down as Chairman to make room for new ideas and fresh momentum,” he said.

His successor is Alfred Vollmer, former Editor-in-Chief of AUTOMOBIL-ELEKTRONIK, whom Hudi wished every success in guiding the future development of the congress.