Abstract
Cyber-physical systems (CPS) for infrastructure monitoring face increasing demands for resilience against faults, cyber-attacks, and environmental disturbances. Digital twins (DTs) offer a promising paradigm to enhance resilience by providing real-time synchronization, predictive analytics, and run-time adaptation. This paper presents a framework for designing resilient CPS using DTs, integrating run-time evolution mechanisms, self-adaptive control, and IoT-enabled sensing. The framework leverages a hierarchical digital twin architecture that mirrors physical infrastructure components and enables continuous reconfiguration to maintain system dependability. A simulation-based case study of a smart bridge monitoring system evaluates resilience metrics including fault detection latency, system recovery time, and operational availability under various disturbance scenarios. Results demonstrate that the proposed DT-driven approach reduces average detection latency by 42% and improves availability by 18% compared to a baseline non-adaptive CPS. Regression analysis further identifies key design parameters that significantly influence resilience outcomes. The study underscores the importance of synchronization frequency and model fidelity in achieving resilient operation. These findings provide practical guidelines for engineers and researchers developing resilient CPS for critical infrastructure.
Keywords
Digital Twin, Cyber-Physical System, Resilience, Infrastructure Monitoring, Run-time Evolution, Self-Adaptation, Internet of Things, Smart Bridge