From RCM to predictive maintenance: The InteGRail approach

Taking advantage of semantically enriched condition information, a new approach in the deployment of maintenance processes becomes possible, which can start a migration from time/mileage based maintenance intervals to maintenance that is prioritised based on condition. This is one of the aims of the InteGRail project, an integrated project partially funded by the European commission and led by UNIFE. The project general aim is to show how better information exchange can bring to increased performance of the railway system, following a holistic approach, which involves all areas: rolling stock, infrastructure, train operation and traffic management. Using ontology based RCM data (see also papers on "ontology driven railway RCM data integration" and "engineering knowledge-based condition analysers for on-board intelligent fault classification: a case study"), the project investigated how it can support existing and new methodologies, particularly in the maintenance field, identifying how existing and new applications can be integrated together for better cooperation, using a standard backbone. A major results is that all applications have been based on a common architecture, enabling easier and cheaper interoperability of information systems. This implied to extend the ontology in order to properly express the maintenance processes and support a number of maintenance applications, e.g. the unplanned event manager and the predictive maintenance server. Some key demonstration scenarios, based on prototypes working in real-life condition, are being implemented in order to prove that the concepts can be successfully implemented and that they can bring the expected results. Final delivery is planned at the end of year 2008.