Evolutionary Computation Modelling for Structural Health Monitoring of Critical Infrastructure

Monitoring of critical infrastructure for Structural Health Monitoring (SHM) is vital for the detection of structural damage (cracks or voids) at an initial stage, thus increasing the structures' serviceable life. The traditional methods of visual inspection to detect damages are time-consuming and less efficient. Sensor based Non-Destructive Techniques (S-NDTs) such as ground-penetration radar, acoustic emission, laser scanning, etc. for detection and analysis are extensively used to monitor structural health but are expensive and time-consuming. Recent advancements in Artificial Intelligence (AI) techniques such as Computer Vision (CV) assisted with Convolutional Neural Network (CNN), Machine Learning (ML) and Deep Learning (DL) in Structural Health Monitoring (SHM) provide more accurate data classification and damage detection systems. This paper provides a state-of-the-art review of the applications of AI-based techniques in SHM. A detailed study on vision data collection, processing techniques, and segmentation (feature, model, and pattern) is discussed, along with their limitations. The application of AI techniques for SHM to detect, isolate, and identify data anomalies, along with biomimetic algorithms are reviewed to assist in future research directions for life critical infrastructure monitoring.

Automated summary

This paper provides a state-of-the-art review of AI-based techniques in Structural Health Monitoring (SHM), which improve the accuracy of data classification and damage detection. It discusses vision data collection, processing techniques, and segmentation methods, along with their limitations. The paper also reviews the application of AI techniques for detecting, isolating, and identifying data anomalies, and the use of biomimetic algorithms. It provides important insights for future research on life-critical infrastructure monitoring.