From data to insight, enhancing structural health monitoring using physics-informed machine learning and advanced data collection methods

Abstract Owing to recent advancements in sensor technology, data mining, Machine Learning (ML) and cloud computation, Structural Health Monitoring (SHM) based on a data-driven approach has gained more popularity interest. The methodology proved be efficient robust compared with traditional physics-based methods. past decade witnessed remarkable progress ML, especially the field of Deep (DL) which are effective many tasks achieved state-of-the-art results various engineering domains. In same manner, DL also revolutionized SHM technology by improving effectiveness efficiency models, as well enhancing safety reliability. To some extent, it paved way for implementing real-world complex civil mechanical infrastructures. However, despite all success, intrinsic limitations such its massive-labelled Requirement, inability generate consistent lack generalizability out-of-sample scenarios. Conversely, SHM, corresponding different state structure is still challenging task. Recent development physics-informed ML methods provided an opportunity resolve these challenges limited-noisy mathematical models integrated through algorithms. This method automatically satisfies physical invariants providing better accuracy improved generalization. manuscript presents sate-of-the-art review prevailing damage inspection, discuss their limitations, explains diverse applications benefits setting. Moreover, latest extraction strategy internet things (IoT) that support present briefly discussed last section.