Scientific paper ID 2564 : 2024/3

The report focuses on the capabilities of predictive maintenance combined with machine learning that can significantly improve the reliability, operability and safety of level crossing systems.

Predictive maintenance uses data collected from sensors and other sources to predict potential failures before they occur. Machine learning algorithms analyze this data to discover patterns and predict when and where maintenance should be performed. Machine learning models can identify subtle changes in the operation of railroad crossings that may indicate impending failure. By addressing problems before they lead to failure, predictive maintenance minimizes unexpected breakdowns. This increases the overall uptime of crossing systems, ensuring that trains and road traffic are less likely to experience delays due to equipment failure. Serviceability refers to the ability of a level crossing system to be in a state ready for use.

Predictive maintenance contributes to safety. By anticipating breakdowns, predictive maintenance can prevent breakdowns caused by equipment failure. Machine learning algorithms can continuously analyze sensor data in real time to ensure that all components of the crossing system are functioning properly.

By using these technologies, rail operators can significantly improve the performance and safety of their level crossing systems, ultimately leading to safer, more reliable and efficient rail services