Integrating condition monitoring in signalling systems

Research Project, 2026

The current railway signalling system employs safety related sensor data (e.g., switch locking data) for management of train operations. There however European plans to increase the scope for the next generation of signalling systems to allow for status monitoring to be incorporated. This would allow a proactive approach to deal with potential causes of traffic disruptions and facilitate planning of (preventive and corrective) maintenance, thus minimising its influence on railway operations.

This integration is pushed by several actors in the sector with different incentives. From a societal perspective it is however vital that the results are efficient and robust solutions. These should e.g., avoid gathering irrelevant data, and unnecessary classification of data as safety related. Flaws in the integration will be costly, may cause lock-in effects, and will remain for a long time. To obtain a scientifically sound implementation must however consider a broad spectrum of aspects related to fields such as railway asset management, large data analysis, and signalling technology.

The project sets out from a (largely existing) broad identification of (vehicle and infrastructure) monitoring data. It evaluates if and how these data are suitable for integration into the signalling system and how integration should be done. It evaluates the potential to employ machine learning in large data assessment. The outcome is a strategy with prioritisation of data to be integrated within a future signalling system and an outline of how the integration should be tailored. The focus is on railways. However, other modes of transportation, e.g., connected road fleets, will see similar challenges soon.

If successful, the project will be extended into a doctoral project with external funding.