Onboard Vibration-Based Condition Monitoring Architecture for Freight Railcar Bearings

Abstract

Bearing degradation produces a fuel and operational penalty invisible to current energy management systems, quantified in a companion paper as over $71 million in excess annual fuel expenditure across Class I consists. Onboard vibration monitoring systems capable of detecting early onset bearing degradation have been developed and tested by the University Transportation Center for Railway Safety and HUM Industrial Technology; however, no architecture currently exists to route this bearing health data into the locomotive's energy management layer for real-time operational optimization. By transmitting the accumulated vibration readings obtained by onboard sensors to the locomotive's main computer, a rolling-resistance adjustment factor can be produced and used to correct the information gap in the optimization software. This paper proposes a fleet-level health data layer in which persistent car-by-car bearing histories accumulate across multiple locomotives and trips, enabling condition-based monitoring and maintenance scheduling, optimized consist assembly and risk-graded cargo assignment. A phased deployment strategy is presented beginning with hazmat transportation and specialty cars where failure consequences far exceed sensor deployment costs. The implications for automated consist management and freight rail data governance are addressed in a companion paper.