Improving the resilience of railway systems

Railway systems are complex systems consisting of an infrastructure network on which a timetable is scheduled for transporting passengers and cargo trains. Trains are operated by rolling stock and crew resources. These are coupled to information and decision-making systems. For smooth functioning of railway systems all these elements need to dynamically interact with each other in a synchronous manner. Due to malfunctioning of one or more of these elements, disruptions in a railway system occur from time to time, which may hinder services for a couple of hours. In particular, due to accumulating disruptions the Dutch railway operations can get into a state of (near) out-of-control, in which barely any traffic is possible. A critical aspect of a railway system is therefore the impact of fluctuations or perturbations on the system’s functionality, which, in the parlance of complexity science, is its resilience. Through modelling railway systems in terms of a complex dynamically-linked layered Network of Networks, we aim to understand disruptions and (near) out-of-control situations, and thereby improve the system’s resilience – i.e., its ability to recover from disruptions – and reliability. Layers include the train service and the decision networks; they interact dynamically via infrastructure, rolling stock and crew allocation and operational information flows. Specific deliverables for the project are early warning signals of out-of-control events, and adaptable control measures and management strategies addressing all layers. We also seek new insights for improving the resilience of wider complex systems that can be considered as Networks of Networks.