![[2602.19614] Workflow-Level Design Principles for Trustworthy GenAI in Automotive System Engineering](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.19614] Workflow-Level Design Principles for Trustworthy GenAI in Automotive System Engineering
Summary
This article presents workflow-level design principles for integrating trustworthy Generative AI in automotive system engineering, addressing challenges in safety-critical applications.
Why It Matters
As Generative AI becomes increasingly integrated into safety-critical systems like automotive engineering, establishing trustworthiness and alignment with verification practices is essential. This research provides actionable principles to enhance the reliability and safety of AI applications in this domain, which is crucial for industry stakeholders and regulators.
Key Takeaways
- Monolithic prompting can overlook critical changes in specifications; a section-wise approach is more effective.
- Requirement deltas can be successfully integrated into SysML v2 models, enhancing traceability.
- Explicit mappings from specifications to architectural components improve regression testing and safety.
Computer Science > Software Engineering arXiv:2602.19614 (cs) [Submitted on 23 Feb 2026] Title:Workflow-Level Design Principles for Trustworthy GenAI in Automotive System Engineering Authors:Chih-Hong Cheng, Brian Hsuan-Cheng Liao, Adam Molin, Hasan Esen View a PDF of the paper titled Workflow-Level Design Principles for Trustworthy GenAI in Automotive System Engineering, by Chih-Hong Cheng and 3 other authors View PDF HTML (experimental) Abstract:The adoption of large language models in safety-critical system engineering is constrained by trustworthiness, traceability, and alignment with established verification practices. We propose workflow-level design principles for trustworthy GenAI integration and demonstrate them in an end-to-end automotive pipeline, from requirement delta identification to SysML v2 architecture update and re-testing. First, we show that monolithic ("big-bang") prompting misses critical changes in large specifications, while section-wise decomposition with diversity sampling and lightweight NLP sanity checks improves completeness and correctness. Then, we propagate requirement deltas into SysML v2 models and validate updates via compilation and static analysis. Additionally, we ensure traceable regression testing by generating test cases through explicit mappings from specification variables to architectural ports and states, providing practical safeguards for GenAI used in safety-critical automotive engineering. Subjects: Software Engineering (cs....