Shaping the future together. With RISC Software GmbH, Materials Center Leoben Forschungs GmbH, eologix sensor technology GmbH and Siemens Mobility we are working on innovations for railway operations.
The FFG research project PRM4.0, a research project of RISC Software GmbH, Materials Center Leoben Forschung GmbH, eologix sensor technology GmbH, Siemens Mobility and voestalpine Railway Systems, has now been in existence for 3 years with the aim of driving innovations and improvements for railroad operations.
The goal of the PRM4.0 project is to create a prototype for a condition-based and predictive monitoring system for railroad infrastructure.
An essential core element is the fusion of track and vehicle data as well as relevant external data sources such as weather data, timetables and dispatch plans. The use of energy-autonomous sensors or virtual sensors, which are currently atypical for the railroad infrastructure or have yet to be developed, and the combination of data- and model-driven analysis methods based on real-time data from a test site will enable accurate forecasts of the short- and medium-term condition of the railroad infrastructure.
The use of machine learning algorithms as well as artificial intelligence methods should enable reliable statements on condition developments of the railroad infrastructure, which will significantly increase availability while reducing life cycle costs.
The main project highlight from the point of view of voestalpine Railway Systems is the detailed analysis of our core product in the form of several railroad turnouts comprehensively equipped with sensors, mapping of the same by means of a digital twin, and validation of the results obtained from this via the measured wear condition. This hybrid approach allows us to generate additional information and thus significant added value for our customers.
Development highlights at Siemens Mobility:
Within the project PRM4.0. the team of RISC Software GmbH is engaged in the development of a Big Data system architecture for the processing of large amounts of data and the analysis of turnout changes.
Specifically, spectral analyses of the sensor data were designed and performed using a Short-Time Fourier Transform, and an evaluation tool was developed that enables semi-automated labeling of the data curves.
The goal is the early detection and derivation of relevant patterns in the data source to determine the wear condition of the turnouts.
A method was developed by MCL to determine the transition point of the railroad wheels from the wing rail to the frog of the turnout using sensor signals. This transition point is an important indicator for the wear condition of the turnout, but also of the wheels, and could previously only be determined during inspections.
This automatic monitoring of the wear condition, which is possible for the first time, opens up new perspectives for maintenance planning.
eologix sensor technology developed a new sensor design in order to be able to install sensors even in limited space on the turnout.
These sensors are triggered externally by light barriers, as soon as a train passes the measurement starts. By measuring different points at the same time, several angles can be simultaneosly observed.
We at voestalpine Railway Systems are looking forward to the remaining project period until autumn of this year and would like to take this opportunity to say thank you to our project partners for the excellent cooperation in the PRM4.0 project.