The shift toward Software-Defined Vehicles is no longer a strategic vision – it is an industrial reality. In this interview, Augustin Friedel, Associated Partner at MHP – A Porsche Company, shares a pragmatic perspective on what will truly define competitiveness in the SDV era as we move into 2026 and beyond.
According to Augustin, the debate is no longer about architectural concepts or middleware choices, but about industrialization at scale. As OEMs ramp up production of software-centric platforms –both electric and conventionally powered – the real challenges emerge: operating models that cannot sustain continuous delivery, fragmented toolchains, supplier ecosystems misaligned with software lifecycles, and legacy platforms that were never designed for continuous evolution.
At the same time, the competitive benchmark is shifting again. While traditional players are still scaling SDV capabilities, new challengers are already building around AI-native vehicle strategies. In this conversation, Augustin outlines where the biggest organizational gaps remain, how companies should prioritize their roadmaps, and why long-term differentiation will depend less on buzzwords and more on industrial capability, control points, and the ability to continuously learn and improve at scale.
The concept of Software Defined Vehicles is evolving rapidly. What are the most significant shifts you see in the SDV landscape as we move into 2026 and beyond?
What’s changing right now is that SDV is moving from theory to industrialization. For years, the industry debated definitions – domain vs. zonal, middleware choices, service-oriented architecture, update strategies. In 2026 and beyond, the decisive factor will no longer be conceptual clarity, but the ability to industrialize SDV at scale. The winners will be those who can manufacture and operate SDV platforms in real volume, not just demonstrate them in prototypes or pilot fleets.
We’re seeing OEMs ramping up production capacity for their next-generation vehicle platforms, that are software-centric by design, independent of whether they are electric or conventionally powered. That shift matters because once you ship real volume, the discussion changes immediately: you discover where your toolchain breaks, where your integration cadence collapses, where your supplier operating model doesn’t match your release model, and where your organization simply can’t keep up with the reality of “software as a live product.”
At the same time, a second shift is becoming unavoidable: Existing combustion engine platforms will live longer than many roadmaps assumed, due to demand volatility, tariff and regional policy changes, and the simple economics of platform lifecycles. The problem is: many of these legacy ICE platforms were never included in the SDV “big bet” programs. So now, one of the top priorities becomes: make the legacy platforms SDV-like by extending OTA capabilities where technically feasible, modernizing E/E constraints pragmatically, selectively introducing virtualization, and upgrading diagnostics and data pipelines. The objective is not to turn these platforms into “perfect” SDVs, but to make them sufficiently SDV-like to enable continuous improvement, reduce variant complexity, and support a modern digital customer experience – until new, SDV-native conventional platforms are ready.
Importantly, this does not contradict the core SDV premise: for newly developed platforms, whether electric or conventionally powered, SDV principles can and should be applied consistently. A newly developed ICE platform that follows SDV principles can be just as capable, scalable, and future proof as an EV platform built on the same architectural foundations.
And third, we’re entering a phase where SDV alone is no longer the endgame. While traditional OEMs continue ramping up SDV programs, challengers from Asia and the US West Coast are already shaping roadmaps around AI-Defined Vehicles. The differentiator of AI-Defined Vehicles is not just software modularity, but learning systems, data loops, and an AI-native product strategy across the full vehicle domain. That changes the competitive benchmark: SDV becomes table stakes, and AI-defined capabilities become the next layer of differentiation.
From an OEM and supplier perspective, where do you see the biggest gaps today between SDV ambition and real-world implementation?
The uncomfortable truth is: Most gaps are not technological, they are organizational.
We love technology debates in automotive. We debate compute, middleware, chip partners, zonal architecture, software platforms and “build vs. buy.” And yes, those topics matter. But what I see repeatedly is that companies try to solve an operating model problem with an architecture diagram.
SDV requires OEMs and suppliers to behave differently: faster decision cycles, clear product ownership, platform governance, DevOps maturity, integrated tool chains, and ruthless prioritization. Instead, many organizations are still optimized for the old world: milestone-driven projects, handovers between departments, procurement incentives that reward cost-down rather than lifecycle value, and siloed domain teams that cannot ship integrated experiences.
I often compare current OEM and supplier organizations to ice cream sandwiches: Layered for stability but lacking end-to-end accountability. While they appear structured from the outside, the internal layers are not connected and fragile under pressure.
SDV challenges this, since your product is never truly “done.” The car is a continuously evolving system. If your organization can’t continuously evolve, the architecture won’t save you.
For suppliers, the gap often shows up as a mismatch between delivery model and customer expectation. OEMs increasingly want continuous integration, frequent updates, software-first collaboration, and shared ownership of quality across the lifecycle. Many suppliers are still set up for “deliver ECU, SOP and change request” logic. That’s not a moral failure; it’s a business model and operating model that must be modernized.
How should automotive companies prioritize their SDV roadmaps over the next few years to balance speed, scalability, and long-term value creation?
If I had to identify one overarching priority, it would be this: transform the organization as aggressively as you transform the product. SDV roadmaps that focus solely on features and architectures will consistently underdeliver. The real shift must happen in the operating model.
The first priority should therefore be the operating model itself – build the factory before you promise the output. This means establishing true software product ownership, end-to-end accountability, and release governance that enables continuous delivery rather than relying on quarterly integration events. If teams cannot ship frequently and safely, the organization does not yet have an SDV capability; it simply has a traditional program under a new label.
Closely connected to this is the need to treat CI/CD pipelines and development environments not as tools, but as strategic assets. Robust CI/CD, automated testing, and standardized environments across teams are not glamorous investments, but they ultimately determine speed and scalability. When teams reinvent pipelines or operate in fragmented environments, integration complexity quickly becomes the dominant cost driver.
Another essential shift is moving toward virtualization and cloud-native development as the default. The industry continues to lose significant time late in development cycles due to hardware-dependent and sequential testing processes. A true shift-left strategy requires extensive simulation, scalable SIL and HIL environments, virtual ECUs, cloud-native tool chains, and early system-level validation. As vehicle complexity increases exponentially, this approach becomes the only sustainable way to maintain pace.
End-to-end toolchain integration must also be treated as a strategic control point. Many organizations still operate with fragmented toolchains, where requirements, software development, validation, and homologation exist in disconnected ecosystems. Without lifecycle-wide traceability, automation, and measurable performance, scaling SDV becomes impossible. Toolchain integration is not merely IT plumbing – it is the nervous system of the SDV enterprise.
Finally, companies must clearly identify and defend their technical and organizational control points. In an SDV context, control points determine long-term leverage: platform interfaces, data access, update orchestration, identity and security layers, app distribution models, customer experience platforms, developer ecosystems, and governance structures. If organizations do not actively define and protect these areas, they risk outsourcing them unintentionally and losing their ability to differentiate.
Balancing speed, scalability, and long-term value ultimately requires speed without chaos, scalability without bureaucracy, and ambition without overengineering. The only sustainable solution is a strong operating model that converts complexity into repeatable, industrialized capability.
What role will partnerships, ecosystems, and cross-industry collaboration play in successfully delivering on the SDV promise?
Delivering Software-Defined Vehicles – and even more so AI-Defined Vehicles – is not feasible without strong partnerships. The scope spans chips, cloud infrastructure, middleware, cybersecurity, developer tooling, AI platforms, maps, connectivity, app ecosystems, service operations, and regulatory frameworks. No single OEM or supplier can lead every layer with world-class capability at the required speed.
Partnerships are essential for three main reasons: speed, cost efficiency, and local relevance. From a speed perspective, companies cannot afford to build every capability internally over extended timeframes. From a cost standpoint, platform economics demand reuse and scale. And from a market perspective, customer expectations, regulatory landscapes, connectivity infrastructure, and service models differ significantly across regions. Partnerships help companies adapt to these variations without dramatically increasing internal complexity.
However, partnerships only create value when managed as a strategic capability rather than a procurement function. This requires clarity about what to own and what to partner. Organizations must distinguish between strategic differentiation and commodity capabilities and define clear architecture boundaries and operating models before entering collaborations.
Execution then demands genuine engineering governance. Successful partnerships rely on shared roadmaps, aligned quality KPIs, integrated tool chains, well-defined interface contracts, and efficient escalation mechanisms. Without this operational rigor, collaboration quickly becomes friction.
Finally, partnerships must be continuously evaluated based on measurable outcomes – speed, defect rates, lifecycle costs, customer impact, and dependency risks. Decisions to strengthen, adjust, replace, or rescope collaborations should be fact-based rather than politically driven.
For both OEMs and suppliers, the core challenge remains the same: collaborate deeply without surrendering future control points. The objective is not to internalize everything, but to own what creates differentiation and partner where scale and shared economics create greater value.
Looking ahead, which technologies or capabilities will become true differentiators in the SDV era – and which are perhaps overhyped today?
One of the most overhyped trends today is the indiscriminate application of AI. “AI-enabled” has become a marketing label rather than a strategic direction. AI will fundamentally reshape products and services, but only if it is deeply integrated into vehicle architecture, data strategies, and operating models. Treating AI as an add-on feature produces impressive demonstrations, but not sustainable differentiation.
True differentiators in the SDV era will be capabilities rather than buzzwords. One major differentiator will be an AI-defined smart cabin embedded within a full-domain AI engine. The cabin is where customers experience value daily, but the real breakthrough occurs when cabin intelligence connects to broader full-domain capabilities such as personalization, intent recognition, multimodal interaction, predictive services, and safety co-pilots – anchored in strong privacy, cybersecurity, and robust compute architectures.
Another decisive capability will be scalable data loops and learning systems. The organizations that win will excel at collecting, governing, and leveraging fleet-wide vehicle data to drive continuous product improvement. This requires industrialized end-to-end data pipelines, systematic labeling and evaluation strategies, and disciplined model lifecycle management.
Software industrialization will also become a key differentiator. The ability to orchestrate updates reliably across diverse vehicle variants while maintaining functional safety, cybersecurity compliance, and customer trust will define competitive advantage.
Compute efficiency and platform modularity will matter more than raw performance metrics such as TOPS. Energy efficiency, thermal management, cost-per-capability, and scalable compute strategies across the vehicle lineup will determine long-term competitiveness. The idea of “one chip to rule them all” is likely to disappoint many organizations once industrialization realities set in.
Another frequently overestimated concept is the idea of purchasing complete reference architectures as a shortcut to differentiation. While reference architectures can serve as valuable starting points or benchmarking tools, believing that a company can source an end-to-end architecture and remain differentiated is naive. Architecture directly influences product evolution speed and control points; it is not a commodity when it defines strategic leverage.
Ultimately, differentiation in the SDV era will not come from the loudest technologies, but from the organizational and technical capabilities that enable companies to ship, learn, and improve faster than competitors – at scale, with quality, and while retaining ownership of what truly matters. That is the practical reality of SDV in 2026 and beyond.