Structural key performance indicators for condition monitoring of concrete bridges using artificial intelligence: a review

Abstract

The actual service life of concrete bridges often deviates from design expectations due to varied loading histories and exposure to environmental and traffic-induced deterioration. Accurately, assessing these deviations is essential; yet, remains challenging. Structural health monitoring (SHM) systems offer a pathway to evaluate bridge conditions and guide maintenance decisions, but their adoption is limited, and the data they generate is often underutilized. Recent advances in artificial intelligence (AI) present new opportunities to enhance SHM by extracting actionable insights from complex datasets. This paper presents a structured review of AI applications in the assessment of real concrete bridges, emphasizing approaches that derive structural key performance indicators. Methods are categorized by input data type and assessed for their performance in damage detection, capacity estimation, multi-type damage classification, signal decomposition, etc. Key challenges are discussed, including data scarcity, interpretability, and robustness. The review concludes with recommendations to advance AI toward practical, scalable implementation in bridge assessment. © The Author(s) 2025. Published by Oxford University Press and Southwest Jiaotong University.