Blockchain-enabled monitoring of public works projects using piezoelectric transducers

Abstract

This study proposes a system for real-time structural health monitoring (SHM) including blockchain technology, deep learning, and piezoelectric transducers. Piezoelectric devices gather mechanical energy from external vibrations and also sense structural changes and transfer them into electrical impulses, acting both as transducers and sensors. principal component analysis (PCA) and root mean square deviation (RMSD) serve to expose structural anomalies in these data. Selected for its capacity to record both past and future temporal dependencies in sequential data and assess these characteristics to predict the health state of infrastructure components, a bidirectional long short-term memory (Bi-LSTM) model. The model exceeds traditional models in high prediction accuracy of up to 98.79% at low voltages with a mean absolute error (MAE) of 1.445, a recall of 100%, and an F1-score of 98.3%. Blockchain technology is used for unchangeable recording of anomaly forecasts and sensor data integrity and automation guarantee. Smart contracts instantly start to offer real-time alarms and distributed system governance from a distance when fault thresholds are exceeded. Especially suited for smart cities, bridges, other essential infrastructure, this design increases openness, reduces dependency on centralized systems. The results demonstrate that this combined approach for intelligent infrastructure monitoring is realistic for sustainable growth.