Artificial Intelligence-Driven Structural Health Monitoring: Challenges, Progress, and Applications

This chapter provides a comprehensive study and critical reflections on the use of artificial intelligence techniques to detect structural deterioration using vibration signals (such as accelerations, displacements, and so on). Machine learning and deep learning-based approaches are seen as promising tools for improving safety and optimizing preventive maintenance plans. However, some authors recognize concerns arising from strictly supervised methods, the “black box” nature of the models and their interpretability by human operators. As a result, the purpose of this work is to give useful information about the current damage detection paradigm, allowing for real-time, non-destructive, and trustworthy predictions regarding construction safety within the context of Industry 4.0. Moreover, issues associated with the application of artificial intelligence for pattern identification and decision-making in monitoring structural anomalies are highlighted and evaluated in recent studies.