Quantum Machine Learning for Infrastructure Risk, Resilience, and Disruption Management: A Review and Research Outlook
Taha Rezaei, Saman Ghaffarian
Quantum machine learning (QML) has gained attention for its potential to leverage quantum advantages in processing high-dimensional data, capturing complex correlations, and exploring large solution spaces, beyond the limits of classical machine learning. In infrastructure risk, resilience, disruption, and recovery analysis, these capabilities position QML as a promising research direction with not yet realised applicability. This study examines whether QML studies move beyond computational novelty to address the substantive requirements of infrastructure-oriented disaster risk management. Using a systematic literature review of records from Scopus and Web of Science, the search retrieved 784 records before deduplication and retained 33 studies after staged screening and full-text eligibility assessment. Studies were evaluated through an infrastructure risk lens. The evidence base is narrow in sectoral scope, dominated by energy and power applications, and more developed in restoration, dynamic security, cyber-physical protection, operational instability, and disruption-oriented operational analysis than in hazard-explicit infrastructure disaster-risk assessment. Hybrid quantum-classical workflows predominate, while many studies rely on simulated, benchmarked, stylised, or case-specific data with limited validation and uneven reporting of data and preprocessing practices. Although selected studies show potential for restoration, classification, anomaly detection, optimisation, and adaptive coordination, the field remains partly connected to core infrastructure-risk concerns including hazard representation, vulnerability, interdependence, cascading disruption, continuity, and recovery dynamics. QML should therefore be understood not as a mature framework for infrastructure-oriented disaster risk management, but as a promising computational frontier with potential to advance the field through broader sectoral application, stronger empirical validation, multi-hazard integration, and benchmarking against classical methods.
Reliability Engineering & System Safety•Aug 3, 2026•0 citations