A Real-Time Framework for Structural Health Monitoring Based on the Internet of Things—An Experimental Study

Abstract

Structural Health Monitoring (SHM) is a continuous and autonomous monitoring technique for ensuring structural safety, integrity, and performance without affecting the structure itself. Advances in technology help in identifying structural defects early on, which aids in making decisions pertaining to when repair and rehabilitation are required for civil structures resulting in saving the economy. The Internet of Things (IoT) and cloud services paradigm extends a faster and more reliable SHM strategy. There have been few studies, especially in data driven SHM for a while, where SHM with IoT was implemented over traditional monitoring techniques. However, solutions presented in the recent literature have the following drawbacks viz, latency, power consumption, and real-time synchronization. Most approaches fail to meet real-time requirements, resulting in slower data visualization, retrieval, and analysis, which might lead to late warning alarms eventually leading to catastrophic events. The present research study focuses on these problems by building a completely feasible, cost-efficient, and scalable IoT architecture utilizing AWS cloud-based services with real-time constraints. Data acquired from the sensor nodes are transmitted via a communication module following star topology to the fog layer (DAQ) through MQTT to the AWS cloud and is processed through different cloud services and is visualized immediately making it an asynchronous architecture. Also, the paper presents the tradeoffs for choosing sensors, communication protocols, algorithms, and cloud services for an efficient and secure real-time approach with an aim to provide a promising platform for an efficient and proactive SHM system for civil infrastructures. Our architecture has been tested with Commercially Off-The- Shelf (COTS) hardware on the prototype models developed in the lab and the results have been validated for the same. The extension of this research work for real-time applications has also been discussed.