IoT-Machine Learning-Based Structural Health Monitoring System for Detection of Cracks in Bridges

Abstract

The deteriorating condition of infrastructure, particularly bridges, poses significant challenges to public safety and necessitates the development of advanced monitoring systems for early detection of structural defects. This paper proposes a novel approach by integrating Internet of Things (IoT) technology with machine learning (ML) algorithms for real-time structural health monitoring, focusing on the detection of cracks in bridges. The proposed system employs a network of IoT sensors strategically deployed on the bridge structure to collect diverse data related to environmental conditions, strain, and vibrations. These sensors provide a continuous stream of data, creating a comprehensive dataset for analysis. Machine learning algorithms, specifically designed for anomaly detection, are applied to this dataset to identify patterns indicative of potential structural issues. This system is designed and implemented using ML and IoT. The excitation results shows that the designed system accuracy and efficiency of crack detection is improved and further enhance overall structural resilience. Furthermore, the integration of a real-time alerting mechanism allows for immediate notification of detected anomalies.