The increasing complexity of software in modern vehicles and the shift toward Software-Defined Vehicles (SDVs) require development approaches that emphasize portability, reusability and platform independence. Traditional Model-Driven Development (MDD) workflows often bind software to platform-specific details early, limiting scalability, reuse and maintainability. This paper introduces PLATO, a framework that integrates Aspect-Oriented Programming (AOP) into MDD, representing the first approach to employ AOP for deferring platform realization, to enable platform-neutral software generation with deferred platform realization. In contrast to existing MDD approaches, PLATO is the first work that defers platform dependencies down to the generated C/C++ code level rather than resolving them at the model level. Functional behavior is captured in a platform-neutral manner and translated into neutral C and C++ code. While manual implementation of core logic is possible, this paper focuses on automated code generation. Platform-specific concerns such as scheduling, communication and diagnostics are implemented as AspectC++ aspects, which are woven into the code at the build of the system. PLATO enforces a clear separation of concerns, improving modularity and maintainability while supporting software reuse across multiple vehicle platforms. A proof-of-concept implementation on a vehicle dynamics Electronic Control Unit (ECU) using a production-grade TriCore compiler demonstrates that PLATO reduces total stack usage by 18% compared to a naïve C-wrapper implementation. Despite this, its performance remains competitive: PLATO runs 1.5% faster than the naïve wrapper and its static code size grows by only 7.5% over the baseline with appropriate compiler and linker optimizations applied. Performance results show that PLATO remains close to a monolithic baseline, where platform logic is embedded directly in the code, while retaining modularity and maintainability. By combining MDD and AOP, PLATO provides a platform-neutral, efficient automotive software suitable for SDVs. The framework supports faster adaptation to new platforms, facilitates software reuse and minimizes resource overhead, offering a practical solution for next-generation vehicle software development.

@inproceedings{paperbot11880,
  title = {PLATO: Platform-Neutral Automotive Software Development leveraging Aspect-Oriented Programming},
  author = {Muhammet Uslu and Atif Imran Penkar and Robert Rasche and Olaf Spinczyk and Michael Unterreiner},
  booktitle = {ICSE},
  year = {2026},
  doi = {10.1145/3774748.3793623}
}