When European Craft Meets Software-Defined Mobility

INTERVIEW Markus Rettstatt, Vice President of Mercedes-Benz Tech Innovation

Software-Defined Vehicles (SDVs) are reshaping the automotive industry far beyond OTA updates.
Markus Rettstatt, Vice President of Mercedes-Benz Tech Innovation, calls this shift a “seismic change in how value is created,” highlighting code-first development, open-source collaboration, AI, Rust, and modular architectures as the key pillars.
More than technology, he stresses the importance of organizational culture and developer experience, calling for shared standards and collaborative ecosystems between Europe and Asia.
This interview with Rettstatt offers a compelling look into how a traditional OEM is navigating its transformation toward a software-driven future.

written by Sang Min Han_Han@autoelectronics.co.kr 

한글로보기





 

The rise of software-defined vehicles is not merely an evolution of OTA capabilities
- it marks a tectonic shift in how value is created in the automotive industry.
To remain at the forefront, Europe and legacy players must rethink their roles across three strategic dimensions: platform innovation, ecosystem agility, and cultural reinvention.    

To begin with, could you please tell us about Mercedes-Benz Tech Innovation and your role and mission within the group?  
Rettstatt        Thank you for having me. Mercedes-Benz Tech Innovation GmbH is the largest German software hub within Mercedes-Benz AG. As a 100 % subsidiary and strategic partner exclusively for Mercedes-Benz we develop technological innovations, digital products, and advanced software solutions. As Vice President Software Defined Car, I lead our efforts in Software-Defined Vehicles (SDVs) and ADAS/Autonomous Driving. I work closely with the Chief Software Officer and Chief Engineer of Mercedes-Benz AG to drive enterprise-wide software transformation - with a particular focus on open-source collaboration for non-differentiating middleware. 


Mercedes-Benz - and indeed all players - are recognizing that SDVs represent not just an evolution in OTA-based services,
but a fundamental transformation of the entire automotive industry structure. In your view, what must Mercedes-Benz, Europe, or the traditional automotive sector achieve to retain leadership in this transition?
Rettstatt        The rise of software-defined vehicles is not merely an evolution of OTA capabilities - it marks a tectonic shift in how value is created in the automotive industry. To remain at the forefront, Europe and legacy players must rethink their roles across three strategic dimensions: platform innovation, ecosystem agility, and cultural reinvention. 
While Chinese EV makers enjoy first-mover advantages, Europe retains formidable strengths - engineering excellence and brand equity among them. Leadership will depend on executing resilient, upgradeable platforms, enabling collaborative ecosystems, and fostering a culture that puts software first. 
We must also harness Europe’s regulatory leadership. For example, streamlining certifications for OTA and cybersecurity across EU markets will accelerate time-to market. Pan-European initiatives - like battery digital twins or V2X infrastructure - are critical. The goal isn’t just to adapt to the SDV era, but to redefine it in a European context - blending cutting-edge software with a legacy of quality, safety, and sustainability.


We often hear that the SDV journey fundamentally starts with semiconductors and the evolution of vehicle E/E architecture.
Could you elaborate on the key issues and challenges related to modular SoCs, chiplets, centralized and zonal architectures?
Rettstatt        The SDV transformation begins with silicon and E/E architecture, but it’s true impact is measured in customer delight, not circuit diagrams. 
Modular SoCs and centralized compute platforms allow vehicles to evolve like smartphones - delivering new autonomy or infotainment features via software updates. 
Zonal architectures reduce wiring complexity, freeing space for immersive displays or sustainable materials. Chiplets and domain controllers optimize performance and cost - but customers care about outcomes: fewer recalls, smoother updates, and smarter in-cabin experiences. 
Crucially, these architectures must support personalization without complication. A centralized platform infused with AI can learn driver behavior - adjusting ride comfort, entertainment, and energy use accordingly. No technical jargon needed, just a car that anticipates. 
To succeed, automakers must abstract this complexity through tight collaboration with chipmakers and cloud providers. Let developers build apps people love, while ensuring robust cybersecurity under the hood. In the end, customers won’t rave about zonal architecture - they’ll rave about cars that feel alive. The silent hero? A tech stack that enables continuous innovation without ever demanding the spotlight. 









Different players seem to emphasize different foundations for SDV - some highlight ‘data-centric’ approaches, others ‘scalability’, ‘customer experience’, or ‘code-first’ development. You’ve been an advocate for architecture-driven development and the “code-first” approach. Why do you believe this principle should take precedence? 
Rettstatt        “Code-First” is a pragmatic enabler for automotive open-source progress. It means prioritizing working, production-grade software - secure protocols, CI/CD pipelines, and real-time telemetry - over speculative standards that remain unimplemented. 
By embracing a “code-first” mindset, we accelerate innovation without compromising safety. The key lies in open governance: creating transparent, inclusive processes that ensure trust, quality assurance, and long-term cross-industry compatibility. This isn’t anti-standard - it’s a faster, more collaborative path to trustworthy, testable standards that actually scale.


From your viewpoint, what does AI mean for SDVs? And how can we safely integrate adaptive AI into mission-critical systems while also contributing to global standardization efforts?   
Rettstatt        AI is a fundamental enabler of SDVs - it powers everything from driver assistance to smart energy management and hyper-personalized user experiences. In this context, AI means adaptability: vehicles that learn, improve, and evolve over time. 
But mission-critical systems demand safety above all, and AI’s probabilistic nature introduces new risks. This requires a layered approach, integrating AI within a robust safety framework and accelerating standards evolution through the same industry-wide collaboration and open governance principles we champion for 'code-first' development. 
Real-time monitoring, deterministic fallbacks, and rigorous testing are essential. Pair AI with traditional controls to guarantee predictability - even in edge cases. At the same time, we must engage with emerging global regulations, from the EU AI Act to ISO efforts, to shape a world where safe, interoperable AI becomes the default, not the exception.


To balance functional safety and development agility, what are the most critical discussions happening around technologies such as POSIX/Linux, Adaptive AUTOSAR, middleware, containerization, or virtualization? 
Rettstatt        Balancing safety and agility in automotive software requires thoughtful engineering choices at every layer. Projects like Eclipse S-CORE offer a compelling vision - an open source core stack for high-performance ECUs that complies with strict safety and security standards. 
Linux remains dominant for infotainment and connectivity, but its non-determinism makes it unsuitable for safety-critical systems. That’s why we isolate safety domains using hypervisors or certified RTOS, while allowing Linux to handle non-critical functions. This hybrid architecture ensures both flexibility and compliance. 
Adaptive AUTOSAR and ASAM have made progress toward openness, but business models and legacy structures often constrain them. The path forward includes modular middleware, containerized workflows, and developer tooling that supports certifiability without sacrificing agility. 
Containerization (e.g., Docker) boosts speed, but safety zones may require stronger isolation - achievable via virtualization. A mixed approach - containers for innovation, VMs for safety - is increasingly common. The industry’s task is to orchestrate both with precision and intent. 


We understand that SDV development requires different languages and evolving standards across domains. What’s driving these trends, and are you already seeing meaningful changes in real-world development environments?  
Rettstatt        The root cause is complexity. SDVs touch everything from brake systems to cloud APIs - each with different reliability, safety, and performance needs. 
Traditionally, C++ dominates safety-critical domains. But Rust is gaining ground for its performance and memory safety. Compile-time guarantees help reduce bugs and enable consolidation of previously fragmented codebases. Centralized computing platforms benefit from Rust’s concurrency tools - enabling real-time orchestration across domains. 
Beyond tools, it’s a mindset shift. Rust’s ecosystem mirrors the collaborative ethos of modern development. Startups embrace this naturally; legacy OEMs must retrain teams, untangle vendor locks, and invest in software-centric culture. The benefits - faster iteration, fewer bugs - are real, and necessary. SDVs won’t wait for perfection.

 
One of the major challenges in the SDV transition seems to be organizational. How do you think we can better support developers - through improvements in tooling, simulation, testing environments, and overall developer experience?
Rettstatt        You’re right. Developers are navigating a maze of legacy systems, mismatched tools, and siloed standards, precisely because the cultural and organizational shift we touched on regarding the adoption of new languages like Rust is still underway. That’s why developer experience must be treated as core architecture. 
We need unified codebases, harmonized toolchains, and consistent APIs. Clear documentation, simulation frameworks, and real-world use cases help reduce friction. 
CI/CD, containerized test environments, and cloud-based dev setups are now essential - not optional. 
We’ve already seen progress: teams adopting DevOps principles, cross-silo collaboration, and modern language stacks. But to scale this transformation, we must embed developer-centric thinking into organizational DNA. Empowerment, not gatekeeping, is the new productivity.



 

Automotive standards need to be living ecosystems
- iterative, testable, and open.
This is how we merge speed with safety and turn standards into enablers rather than roadblocks.

SDV emphasizes openness and collaboration. As the platform expands, so do potential connections with adjacent industries. Why is open-source collaboration so important in this space, and how is Mercedes-Benz navigating the balance between open contribution and commercial differentiation?
Rettstatt        Open-source is strategic infrastructure - it’s like building shared highways so everyone can drive faster. Customers don’t care about proprietary middleware - they care about luxury, safety, and innovation. 
By collaborating on non-differentiating stacks, we can focus resources where it matters: on experiences that excite. SDVs interact with smart cities, wearables, and energy grids. 
If each OEM builds in isolation, fragmentation kills progress. Openness ensures interoperability, while differentiation defines brand value. 
On behalf of Mercedes-Benz, we actively contribute to initiatives like Eclipse SDV, because shared platforms unlock speed, reliability, and trust. This frees our teams to focus on bespoke UI/UX, AI-enhanced autonomy, and the engineering of desire. Open isn’t charity - it’s a catalyst.









There’s a well-known saying that “regulation defines the automotive industry.” In fact, regulation, standardization, and technical leadership often go hand in hand. You once mentioned that “traditional standardization is too slow for SDVs.” Could you explain what you meant by that, and how we can create standards that match the pace of modern software-driven development?     
Rettstatt        The challenge is one of rhythm. Hardware development runs in decades; software evolves in weeks. Yet we still treat software standardization like its hardware. 
Take OTA updates - once exotic, now expected monthly. If the standards process takes years, it can’t keep up. Instead, as I've advocated, the "code-first" principle should extend to how we develop standards: we must build working code collaboratively and let the standards emerge from practice, not just theory. 
Upstream-first collaboration is vital. If a supplier modifies shared code, they must contribute back - preventing fragmentation. We should define modular APIs that encourage reuse but allow brand-specific layers - like Android vs. Pixel or Samsung skins. 
And we must modernize governance. Think GitHub workflows, not static PDFs. 
Automotive standards need to be living ecosystems - iterative, testable, and open. This is how we merge speed with safety and turn standards into enablers rather than roadblocks.


Finally, do you have any message for developers or policy leaders in Korea, Japan, and the wider Asian automotive ecosystem who are actively thinking about their SDV strategies?
Rettstatt        The SDV transformation is a shared opportunity to co-create the future of mobility. Asia leads in semiconductors, consumer electronics, and manufacturing excellence. By leveraging this strength in open collaboration, we can jointly develop platforms that are safe, adaptive, and globally interoperable. 
We’re already working through initiatives like Eclipse SDV to build open, modular software foundations. We welcome even deeper collaboration with Korean and Japanese innovators - your strengths in user-centric design and precision engineering are essential to shaping this era. 
We support initiatives that enable this future by investing in shared standards, privacy frameworks, and digital talent. SDVs are not a solo journey - they’re a collective leap. 
Let’s build the road together.

---