Fiber-Optic Sensor-Based Structural Health Monitoring with Machine Learning: A Task-Oriented and Cross-Domain Review

SHM is becoming important for the ongoing assessment of aging infrastructure that has safety concerns and is becoming challenged by changing environments and operational needs. For some time now, fiber-optic sensors (FOSs) have been a focus of concern for SHM because of the versatile and durable nature of FOSs for measurements in harsh environments and for serving the diverse needs of multiplexing in both local and fully distributed measurement applications. Additionally, the rapid development of machine learning (ML) makes it possible to obtain insights from very large and complex sensing data sets and to apply novel techniques to do so. This paper presents the first systematic review of the SHM

systems that use FOS and ML and addresses FOS and ML use in SHM for various types of infrastructure, including aerospace. The paper critically reviews fiber optic technologies for measurement, cognitively dividing them into types of systems such as point-based, quasi-distributed, and fully distributed and reviews the effectiveness of each of these technology types for measuring strain, temperature, and structural systems’ spatiotemporal responses. Additionally, the paper examines ML processes for tasks such as damage detection, localization, and severity assessment, and also for the role of the environment in the damage assessment and prognosis of structural systems, focusing on the alignment of sensing systems with appropriate ML paradigms. This helps to

answer the concerns regarding the development of structural systems that have assessment capabilities for robust integrity evaluations.

Message: Hi FE, I need help with “Fiber-Optic Sensor-Based Structural Health Monitoring with Machine Learning: A Task-Oriented and Cross-Domain Review” in “Aerospace Engineering”