A centrifugal mechanism hybrid energy harvester with AI-enabled self-sensing and self-powered for railway monitoring

Abstract Wireless sensor networks (WSNs) and artificial intelligence (AI) are vital for improving the safety and efficiency of railway transportation systems. Therefore, innovative and sustainable power solutions are necessary to address the limited energy resources of WSNs. This study tackled the limited energy of WSNs by developing the Centrifugal Spring and Magnetic Linkage Harvester (CSMLH), a self-powered, AI-enabled hybrid energy harvester for smart railway monitoring. The CSMLH bridged the gap in the field by integrating three energy harvesting mechanisms: solar (SC), electromagnetic (EMG), and piezoelectric energy harvester (PEH), into a single compact system. The main innovation is the bio-inspired, frictionless wind turbine with a centrifugal spring mechanism that dynamically adjusts magnet spacing to optimize magnetic coupling and EMG and PEH power output across a wide range of wind speeds. Additionally, the system incorporates a deep learning algorithm via the OpenBCI platform to process PEH and EMG voltage signals, enabling AI-enhanced self-sensing and real-time data analysis. The experimental results demonstrated the system’s strong performance. The EMG component produced an average power output of 5.6 mW, while the PEH harvesters yielded up to 14.45 mW for the upper PEH and 13 mW for the lower PEH. The solar component proved scalable, with four panels generating 9 W of output power. The integrated AI model achieved high recognition accuracy, with peak training and validation accuracies of 98.85% and 99.45% for PEH signals, and 99.5% and 99.65% for EMG signals, respectively. This high-accuracy AI component marks a significant step toward fully autonomous railway infrastructure.