This is the authors' abstract. We don't add key points for this paper.
Structural Health Monitoring (SHM) has become an important component of modern civil engineering because infrastructure must be assessed continuously throughout its service life rather than only through periodic visual inspections. Recent developments in artificial intelligence (AI), machine learning (ML), deep learning (DL), computer vision, smart sensors, Internet of Things (IoT) platforms and digital twins are changing the way structural data are collected, interpreted and converted into maintenance decisions. This review examines the principal methods and applications of AI-driven SHM for buildings, bridges and other civil infrastructure. It discusses vibration-based monitoring, vision-based inspection, fibre-optic and other smart sensors, anomaly detection, damage classification, remaining-life assessment, predictive maintenance and multi-modal data fusion. The review also considers the roles of convolutional neural networks, recurrent neural networks, autoencoders, ensemble learning, transfer learning and physics-informed approaches. Recent literature indicates that AI can support automated damage detection, structural response estimation and large-scale data analysis, but reported performance is often dependent on controlled datasets and study-specific conditions. Major barriers to field deployment include limited labelled data, environmental and operational variability, class imbalance, sensor faults, model interpretability, uncertainty, transferability and cybersecurity. The paper argues that future AI-driven SHM should move toward physics-informed, explainable and uncertainty-aware systems integrated with IoT, edge computing and digital twins. Such integration can support continuous condition assessment and more proactive infrastructure management while retaining engineering judgement as an essential component of safety decisions.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Nilofar Shaibaj Patel (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: