This is the authors' abstract. We don't add key points for this paper.
Certification of aircraft systems is a complex task that is difficult to automate requiring significant subjective decision making. Established certification standards such as CFR-25 and MIL-HDBK-516 require considerable subjective analysis to transform the certification requirements into meaningful actionable requirements. This prevents the automation of any verification, assurance and certification task. While established methods rely on automation of assurance case generation, several key tasks that still requires human input preventing the co-creation of designs, verification scenarios, evidences, and assurance cases. Building on the recent success of large language models, we develop a framework that enables the automated verification and assurance of cyber-physical systems. Our framework consists of two parts – (i) an automated pipeline permits the automatic extraction of information from system artifacts such that a they can be semantically linked using a graphical representation; and (ii) an automated pipeline that enables the synthesis of verification scenarios to co-generate evidence along with the assurance case of the cyber-physical systems. We demonstrate our framework on a concrete example of a landing gear sub-system of the aircraft and highlight the benefits that can be realized through the automation of the bottlenecks in the task of verification, assurance, and certification. Using semantically linked representations of the knowledge, we enable complex reasoning on the knowledge contained in system artifacts providing meaningful feedback to the designers and certifiers on means improve the overall system. Our investigations on the landing gear use-case demonstrates the feasibility of the use of large language models to support the systems engineering tasks for cyber-physical systems and the use of a knowledge graphs in the construction and assessment of the cyber-physical system's assurance.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Jin et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: