VOLTAGE-BASED STRUCTURAL HEALTH MONITORING USING IOT AND MACHINE LEARNING: A FAULT DETECTION APPROACH

Voltage-Based Structural Health Monitoring System is an IoT- and AI-powered solution designed to monitor and detect anomalies in electrical voltage patterns within critical infrastructure such as buildings and bridges. The system utilizes the ZMPT101B voltage sensor to continuously capture real-time AC voltage data, which is processed through an ESP8266 microcontroller. This data is then transmitted wirelessly to the cloud using platforms like Blynk. Machine learning models—LSTM, Random Forest, and Decision Tree—are trained to classify voltage behavior as normal or fault-prone. The models are evaluated using metrics such as accuracy, precision, F1-score, and RMSE. A user-friendly Streamlit dashboard presents real-time visualizations, model comparisons, and fault detection results, making the system accessible for both technical and non-technical users. The project emphasizes data preprocessing techniques such as normalization, noise filtering, and synthetic data augmentation to improve model performance. This modular, low-cost, and scalable solution enables early detection of electrical faults, reducing risks associated with infrastructure failure and ensuring timely maintenance. By combining IoT with AI, this system enhances safety, reliability, and efficiency in structural monitoring applications.