The Underestimated Complexity of Software Transformation

Becoming a “softwarefirst company” is not a matter of awareness – it is a question of execution. Many German car manufacturers face a tough reality: millions of lines of legacy code, hundreds of proprietary interfaces, and IT architectures that envolved over decades and became slow in the process.
Software development follows different rules than traditional automotive engineering. While local OEMs invested billions, Chinese competitors worked from a clean‑sheet approach – leveraging modern tech stacks, cloudnative‑ architectures, and agile development – without AUTOSAR slow downs.

Our experience: we help OEMs untangle this complexity – from identifying technical debt to implementing modern development practices. The key lever is architecture decisions that sustainably improve scalability and time‑to‑market. This can be described as a three‑horizon architecture strategy:

• Horizon 1: Immediate Stabilization (“Legacy Containment”)
  Critical legacy systems are stabilized using API abstraction layers. Monolithic ECU software is wrapped into microserviceready modules without touching the core system.
  Result: faster feature delivery with minimal risk.
  
• Horizon 2: CloudNative Development Environments
  In parallel, establish cloudnative environments with DevSecOps pipelines. Shift from component teams to crossfunctional feature teams.
  We’ve supported OEMs to cut feedback cycles from months to weeks by adopting “you build it, you run it.”
  
• Horizon 3: Strategic Target Architecture
  In the long term, an architecture based on consistent domaindriven design emerges. This clearly separates:
  Commodity code (platform services such as authentication, telemetry)
  Differentiation code (e.g. UX, driving dynamics algorithms)
  So resources flow where they create real competitive advantage.

The Platform Dilemma: Build vs. Buy

Software‑defined vehicles decouple hardware and software. Platforms such as Nvidia Drive, Qualcomm Snapdragon Ride, or Android Automotive enable modular development.
The strategic question remains: Should an OEM really try to build the next Android or iOS? Or is it smarter to use existing platforms and focus fully on a distinctive UX layer?
The industry shows two successful, yet very different paths:

• Tesla: full integration of the entire stack
• BYD: a pragmatic best‑of‑breed approach

Our approach: support OEMs in making this decision based on facts, not ideology. The key lever is a clear definition of where differentiation truly matters: where in‑house development pays off, and where integration is faster and more cost‑effective.

This leads to an evaluation framework we call Strategic Technology Value Assessment (STVA), based on three dimensions:

• Technological differentiation: Where do we create clear customer value?
• Control and dependency risk: How do we quantify platform lock‑in costs over five‑year cycles?
• Time-to-competence: How long will it realistically take to reach competitiveness?

The Overlooked Factor: Data and AI

Autonomous driving, predictive maintenance, personalized services – none of this comes from software alone. It requires powerful data pipelines and AI infrastructure.
Chinese OEMs have a clear advantage here. They collect far more vehicle data, connect their ecosystems seamlessly with super‑apps like WeChat, and generate a strong data flywheel effect: more data → better AI → better products → more users → even more data.

German OEMs need to catch up under stricter regulation, which makes the gap exponentially harder to close.

Our experience: a central lever is making the value of data more visible – for OEMs and customers. When customers clearly benefit from datadriven services, acceptance increases.
For example, we develop applications that improve charging interoperability by consolidating live availability data, harmonizing tariffs and roaming information, and standardizing APIs – especially where today heterogeneous protocols, incompatible backend systems, and missing end-to-end data flows cause inconsistent status messages and uncomfortable charging experiences.

The Ecosystem Play

We add one crucial dimension to the usual scenarios: OEMs as orchestrators of automotive ecosystems. Instead of building everything in-house, the focus shifts to:

• Proprietary vehicle platforms – hardware excellence remains core
• Opensource software platforms for third parties
• Premium integration and user experience as differentiators
• Control of the customer interface and data

Our experience: the mindset shift is from “Not Invented Here” to “Best Integrated Here.” This will decide whether OEMs become leaders – or just feature suppliers inside someone else’s platform.

The Time Factor: Why Speed Is Everything

Large‑scale software transformations realistically take 7 to 10 years. At the same time, competitors operate in 2‑ to 3‑year cycles. That means: Even if German OEMs do everything right today, the gap will initially widen – this is mathematically unavoidable.
The challenge even intensifies by an additional factor: the core business must fund the transformation while it‘s profitability declines. This double bind makes speed not just important – but existential.

The challenge even intensifies by an additional factor: the core business must fund the transformation while it‘s profitability declines. This double bind makes speed not just important – but existential.

Our Conclusion

This transformation is not just an engineering challenge. It is an IT, data, and cultural revolution that will decide who remains relevant in the next decade. Winners will be OEMs that master complex technology and data ecosystems – not just those that build excellent cars.
The resources exist. The real question is: is there enough courage to decide fast enough?

As a partner for digital transformation in automotive, we support OEMs across the key levers – from software architecture and data strategy to AI enablement. The next 3 to 5 years will decide the field. We bring architecture, tooling, and delivery experience that measurably increase speed and quality.