A Verification Framework for Certifying Learning-Based Safety-Critical Aviation Systems
This work addresses safety certification for aviation systems using AI, which is critical for regulatory compliance and public trust, though it appears incremental by combining existing verification approaches.
The authors tackled the problem of ensuring safety in learning-based aviation systems by developing a verification framework that integrates offline mixed-fidelity tools and online monitoring methods, resulting in a loosely coupled design that supports continuous safety assessment throughout development and deployment.
We present a safety verification framework for design-time and run-time assurance of learning-based components in aviation systems. Our proposed framework integrates two novel methodologies. From the design-time assurance perspective, we propose offline mixed-fidelity verification tools that incorporate knowledge from different levels of granularity in simulated environments. From the run-time assurance perspective, we propose reachability- and statistics-based online monitoring and safety guards for a learning-based decision-making model to complement the offline verification methods. This framework is designed to be loosely coupled among modules, allowing the individual modules to be developed using independent methodologies and techniques, under varying circumstances and with different tool access. The proposed framework offers feasible solutions for meeting system safety requirements at different stages throughout the system development and deployment cycle, enabling the continuous learning and assessment of the system product.