This paper explores the concept and implementation of a diagnostic system leveraging Service-Oriented Vehicle Diagnostics (SOVD) within the context of Automotive Service-Oriented Architecture (ASOA). A key challenge in Software-Defined Vehicle (SDV) architectures is reduced traceability and non-transparent data flow due to dynamic system setup, which hinders the detection and understanding of potential software errors. Whereas for service-oriented architectures in domains like cloud computing, error traceability is already state-of-the-art, in SDVs it is still in its infancy. We seek to contextualize diagnostic data to support root cause analysis and facilitate predictive maintenance. We demonstrate an approach that provides mechanisms for effect chain tracing as well as degradation detection. Through the use of a digital twin of the SDV architecture, integrated postmortem analysis and scenario-agnostic error replay become possible. The case study shows that the system can visualize and analyze ASOA traffic, which facilitates the diagnosis and troubleshooting of modern automotive systems.
Molz et al. (Thu,) studied this question.