Structural health monitoring (SHM) is a critical research topic in civil engineering for assessing the integrity of constructed facilities, yet its widespread deployment is often hindered by the high cost of commercial equipment. This study introduces an accessible, vibration-based SHM system consisting of a slave unit for data acquisition via an MPU6050 sensor and a master unit for long-range wireless transmission using the LoRa protocol. To overcome the inherent noise levels of inexpensive MEMS sensors, we propose a robust modal identification framework that utilizes the Levenberg–Marquardt optimization method combined with a sliding window strategy to accurately estimate damped natural frequencies. Experimental validation conducted on a steel beam demonstrates the technical viability of this event-triggered IoT architecture. The designed system achieved a relative error of only 6.38% in natural frequency identification compared to a high-precision commercial reference system. Ultimately, this framework provides a technically sound, resource-efficient solution for structural assessment.
Vu et al. (Sat,) studied this question.