Improved condition assessment through statistical analysis
The railway is composed of more than rails, sleepers and overhead wires; if maintenance stops, soon all over railway operations will stop too. Railway maintenance should be done where the benefits are maximized.
To maximize the benefits and reduce system risks, maintenance should be performed before things fail, rather than as repairs when things are broken. Preventive maintenance can be time based or linked, for example, to the load a railway section has been subjected to. Even better is to perform maintenance based on the state railway is in. The very best situation would be if we can create forecasts of when a property will have deteriorated so much that maintenance is needed far in advance, so that we can plan maintenance actions. Such maintenance may then be done without need to cancel trains or creating delays.
In this project we study if collected railway condition data can be used to improve the condition assessment by studying the condition changes from measurement to measurement. Today, only the latest condition measurement is used and the major project hypothesis is that the information buried in older measurements can be used more efficiently.
The researchers in the project are Bjarne Bergquist and Peter Söderholm. The project includes, in addition to Luleå University of Technology also The Swedish Transport Administration, LKAB Malmtrafik and Infranord. The project began October 1, 2013 and runs until 31 December 2016. It is funded by Luleå Railway Research Center and The Swedish Railway Administration
Updated: