Automated Assessment of Complex System Vulnerabilities via Hyperdimensional Network Analysis & Causal Disruption Modelling

**Abstract:** This research introduces a novel framework, Hyperdimensional Vulnerability Assessment and Resilience Enhancement (H-VARE), for proactively identifying and mitigating vulnerabilities within complex, interconnected systems, focusing on fan control systems within large-scale industrial facilities. H-VARE leverages hyperdimensional computing (HDC) for efficient representation and analysis of system state, combined with causal disruption modelling (CDM) to predict the impact of targeted interventions. The system distinguishes itself by moving beyond reactive vulnerability scanning towards a proactive, dynamic assessment of resilience, providing actionable insights for strengthening system defenses against both accidental failures and malicious attacks. We demonstrate the feasibility of H-VARE through simulations on a simplified model of an industrial ventilation fan network, showcasing its potential for significantly improving system robustness.