Academics4Rail

To develop algorithms for automatic detection of anomalies and analytic tools for predicting the degradation of railway asset and facilitate the development of predictive maintenance. To develop decision support framework and tools for integrated Prognostics and Health management for asset maintenance planning and scheduling.

Partners WP8

Luleå University of Technology (LTU), University of Applied Sciences and Arts of Southern Switzerland, Slovenian National Building and Civil Engineering Institute, Vilnius Tech, University of Žilina, Brno University of Technology, University of Huddersfield, Trafikverket.

Goal

Academics4Rail is a scientific community that in an organized way can share and exchange scientific knowledge with ERJU and ERRAC. Academics4Rail is led by EURNEX with 21 European universities connected. The scientific community engages in specific themes creating 6 PhD positions that will enlarge the knowledge in 6 specific areas and will enable the cooperation of academia with industry. LTU are leading WP8 i.e., PhD5 in Prognostics and health management approach for railway asset maintenance. The goals are to a) develop algorithms for automatic detection of anomalies and analytic tools for predicting the degradation of railway asset and facilitate the development of predictive maintenance, and b) develop decision support framework and tools for integrated prognostics and health management for asset maintenance planning and scheduling.

Project status and results

The PhD candidate will be supervised by the host Principal Investigator at LTU, and co-supervised by other partner universities. The candidate will have regular meetings with Trafikverket experts for data collection, data pre-processing, analysis of results, and validation of the system. The candidate will visit other stimulating and creative environments, advanced training, tailored career development and mentoring programs, and access to high-end technologies and facilities from other institutions.

The main objective of this work package is to develop a decision support framework using prognostics and health management approach for railway asset maintenance. These objectives can be divided into the following sub-objectives:

• To identify the specifications and requirements of use case
• To perform data mining, cleaning, pre-processing, etc., for exploiting the data
• To develop anomaly detection models using machine learning
• To develop hybrid models for prediction of asset health
• To implement and develop an integrated framework for maintenance decision support for PHM concepts
• To promote the transfer of knowledge between the project’s participants and to disseminate, fully communicate and exploit the research outputs.