Autonomy Vulnerability Analysis using System Taxonomy (AVAST)

The Department of Defense (DoD) is accelerating the development of autonomous vehicle sys-tems for complex air combat missions, highlighting the need to identify and mitigate potential vulnerabilities. SSCI proposes the Autonomy Vulnerability Analysis using System Taxonomy (AVAST) framework to address this critical requirement for the Navy. AVAST aims to develop a comprehensive ontology of vulnerabilities within autonomous systems, utilizing Model-Based Systems Engineering (MBSE) and expert insights to identify and analyze threats. This frame-work will assess risk levels, pinpoint attack vectors, and create both exploitation and mitigation strategies, providing a combinatorial view of system vulnerabilities that consider the interac-tions and dependencies between different components.In Phase I, AVAST will focus on creating a detailed model of the autonomous system, identify-ing vulnerabilities through Model-Based Vulnerability Analysis (MBVA). This includes em-ploying threat modeling, attack trees, and simulation-based security modeling. The approach will integrate digital twin and digital thread development, ensuring a thorough traceability of vulnerabilities across the system's lifecycle. Combinatorial optimization techniques will be used to discover complex attack sequences, leveraging tools like OptaPlanner to convert MBSE ob-jects into constraint satisfaction problems. The results will be documented and presented in both human-readable and computer-readable formats, facilitating comprehensive stakeholder under-standing and automated cybersecurity analysis.Looking ahead, Phase II will shift to red-team analysis, exploiting the identified vulnerabilities to understand potential attack sequences and their impact. SSCI will integrate MBSE with simu-lation tools to model these sequences accurately, employing a hypergame approach to simulate adaptive tactics and countermeasures. This phase will provide critical insights into the weak-nesses of autonomous systems, enabling the development of robust defensive strategies. The AVAST framework's systematic and innovative approach will enhance the resilience and effec-tiveness of autonomous vehicle systems in complex combat environments, ultimately strength-ening the Navy's operational capabilities.