Smart Autonomous Diagnostics of Switches and Crossings

This paper shows the results of the Turnout 4.0 applied research project.The project focused on the development of an autonomous diagnostic system.A modular diagnostic system is described, including a sensor part installed on the crossing and the vehicle, data acquisition and storage in database systems, evaluation of signals by two systems -Train Identification System and Diagnostic Switches & Crossingsusing artificial intelligence algorithms, Machine Learning and Bayesian statistical analysis.The extension of the system is the determination of the final score, describing the technical condition of the crossing, evaluating the score development over time, and displaying the monitoring results to the infrastructure manager.