Fiber-optic sensor-based proactive railway track monitoring using vibration analysis and neural networks

Abstract

The safety and efficiency of railway infrastructure directly depend on the integrity of the track infrastructure. However, transitional monitoring methods such as visual inspections, non-destructive testing, and track circuits are mostly reactive or periodic. They leave significant time intervals during which critical defects can appear and develop, while their effectiveness is limited by their scope, high cost, and vulnerability to external factors. This creates an urgent need for innovative, automated systems that are capable of providing continuous, high-precision, and most importantly, proactive monitoring of railway track conditions in real-time.