Railway infrastructure monitoring framework (RIMF) and its application will enable predictive maintenance to improve resilience, preparedness and proposed early warning system for the Canadian railway transportation network

Through aggregated real-time live rail track geometry data collected by our rail track sensing units networks, Satellites and environment data models, predictive maintenance algorithms continuously analyzes the condition of rail and rail track during normal operation to reduce the unexpected incidents and failures. 

Predictive maintenance also informs on overall infrastructure resilience and supply chain performance. Predictive maintenance benefits railroads in many areas, including reductions in accidents, unnecessary preventative maintenance, fleet size, spare parts stock, and excess line capacity buffer. Predictive Maintenance relies on advanced statistical methods - machine learning, to dynamically define the track condition. It searches patterns across all sensors to build one multivariate prediction model. The more data sources and data available the better are the predictions. Maintenance intervention is organized based on the predicted probability of component failure in future.