Digital Twins and AI in Infrastructure Engineering: A Global Review of Risk-Informed Design, Operations, and Maintenance

As global infrastructure systems become increasingly complex and vulnerable, the integration of Digital Twins (DT) and Artificial Intelligence (AI) has emerged as a transformative strategy for risk-informed design, operations, and maintenance. However, a significant research gap remains in understanding the global patterns, integration challenges, and practical impact of DT-AI systems across different infrastructure sectors. This systematic review, guided by the PRISMA framework, synthesizes findings from 126 peer-reviewed studies sourced from Scopus, Web of Science, and IEEE Xplore. Analysis revealed that 68% of implementations focus on predictive maintenance and real-time monitoring, while only 12% address early-stage design optimization highlighting an imbalance in lifecycle focus. Furthermore, projects that applied AI-enhanced DTs achieved up to 30% reduction in unplanned maintenance events and improved infrastructure lifespan predictions by an average of 22%. Case studies from Singapore, the UK, Norway, and the US demonstrate real-world benefits in city planning, structural health monitoring, and transportation. Despite these successes, key barriers persist, including data interoperability, cybersecurity vulnerabilities, high implementation costs, and insufficient regulatory standards. This review underscores the need for cross-sectoral collaboration, global policy frameworks, and inclusive innovation strategies to fully leverage DT-AI capabilities in building resilient, adaptive infrastructure.