Design of a Robust Embedded Sensor Network for Real-Time Hull Structural Health Monitoring in Marine Vessels

Abstract

This research introduces a robust sensor network that is embedded in the hull of a marine vessel for real-time structural health monitoring. The system involves strain sensors, an (MEMS) accelerometer, and the wireless communication modules for securing data of structural behavior under different conditions. The most important part to take into consideration is that the selection of the spots like the middle part of the vessel's bottom plate, side shell frame, and joint between the stiffeners as the position of the sensors together with the associated signal conditioning. The signal conditioning involved the Butterworth filter to support high signal-to-noise ratio in improving data quality, by about 15 to 17 dB. The communication framework ensures the reliability of the system with a packet reception rate of 94.87 percent and 66.15 milliseconds of mean latency under dynamic conditions. An embedded, rule-based alert system attains an accuracy of 76.69% for real-time anomaly detection. Also, 80.43% accuracy can be accomplished by a data-driven diagnostic approach with the machine learning model. In general, the proposed framework demonstrates an achievable and scalable approach for rehabilitation of marine structural safety and monitoring efficiency.