Project

TRAINS - Sub-project UV12: Digital methods for predictive maintenance of rail vehicles

Fatigue phenomena of components can be assessed using fatigue strength tests. However, such tests also have errors, meaning that damage can never be completely ruled out. Rather, the verification provides information on the statistical probability of damage. However, this is problematic for assessment in the vehicle sector, as the calculation of the operating times of technical systems is intended to ensure operational safety and extend operating times. The reliability of the predictive methods is therefore of particular importance, especially for the drive train of rail vehicles. In terms of maintenance, this has a particularly high potential for optimization. The main necessity of maintenance for drivetrains results from the highly dynamic, oscillating load on all components, as a consequence of which fatigue damage can occur.

The overall aim of this joint project TRAINS - sub-project UV12 is to introduce the digital methods of Industry 4.0, big data, Internet of Things and cloud technologies as well as data evaluation using machine learning methods into rail vehicle maintenance. The aim is to determine predictive maintenance intervals in order to maximize both the reliability and service life of the vehicles - which in turn should minimize maintenance and operating costs.

The sub-project is being carried out by a total of seven partners who are working on the introduction of "health monitoring" in railcars as well as the establishment of machine learning for data evaluation.

© AG Griehl