INTERNET OF THINGS APPLICATION IN STRUCTURAL HEALTH MONITORING

Abstract

Structural Health Monitoring (SHM) plays an important role in predicting accidents and facilitating incident management. Internet of Things (IoT) provides remote access to sensing data. However, like most monitoring systems deployed in real life, human involvement is still required to examine the structural condition, which can be cumbersome and impractical. To address this problem, a convolutional neural network model is trained and deployed on Raspberry Pi to perform real-time surface crack detection. The crack detection model is then integrated with an IoT monitoring system to tackle both monitoring and inspection processes of SHM. When overloading is detected, the crack detection model will be activated automatically to perform a surface scan. The real-time results are saved in an SQLite database and are updated automatically on a website page. An alert email will be sent if the model detects any crack. Besides, a GUI is designed increase user-friendliness.