A heuristic approach for predictive diagnosis of wheel wear based on low-cost track-side equipment

Predictive diagnosis of wheel wear plays a fundamental role in maintenance of railway vehicles. While on-board installation of wear sensors is affordable only for modern high-performance trains (e.g., high-speed trains), evaluation of wheel wear is mostly performed through periodic measures in maintenance sites for low-performance trains (e.g., secondary line and freight trains). To support constant monitoring of wheel wear at low cost, we present an economical wayside system able to estimate the flange height of wheels, which actually increases with the wheel wear. The approach exploits the capabilities of an innovative electromagnetic treadle and a heuristic estimation method. Feasibility and performance of the proposed approach are evaluated through laboratory and on-field experiments in the context of the INDIGO project.