“Dutch train passengers will benefit tremendously if we can solve this problem”

When Nederlandse Spoorwegen (NS, the Dutch Railways) expressed that they were looking for quicker recovery solutions following railway disruptions, Deb Panja took the challenge. Together with a team of researchers from Utrecht University, Erasmus University Rotterdam, TU Delft and TNO, Panja has landed an NWO Complexity in Transport and Logistics grant worth €590,000, of which €90,000 is co-funded by NS and ProRail. The team will investigate disruptions and (near) out-of-control situations in the Dutch railway system, and work to improve the system’s resilience and reliability. The research team will seek to develop a complex systems approach, in particular modelling the rail network as a network of networks. Interestingly, similar approaches have sometimes been used in climate and evolutionary ecology research.

The project was originally conceived after Panja and Stefan Dekker (Copernicus Institute) heard a pitch by NS at the launch of the National Platform of Complex Systems. Panja explains: “The problem they pitched resonated very well with us. It appeared to be a classic opportunity to showcase how fundamental research could address a problem of high impact on public life. That evening, when I got home and excitedly told my wife about the conversation we had with NS, she responded: ‘It would have a major impact on everyday public life if you can solve this problem.’ I was moved by that remark.”

Collaboration

In December 2015, Panja and Dekker approached Henk Dijkstra, Professor at the Institute for Marine and Atmospheric Research Utrecht (IMAU), to harness his expertise on networks. With a small workshop grant from NWO, they invited researchers from Utrecht, Rotterdam, Delft, NS and ProRail, and TNO in March to brainstorm on the problem and write a proposal, which was funded in November 2016. Panja is really happy with the collaboration from ProRail and NS. “I really hope that shortly after the research is concluded, NS will be able to announce that their disruption handling is working much better than before.”

Bird flocks, stock markets and trains

“A complex systems approach to modelling and simulation resides at the heart of my work,” says Panja. “I focus on stochasticity and on materials. A hallmark of a complex system is that it consists of a large collection of interacting individual elements, from molecules in a material to interacting species in an ecosystem and individuals in a crowd. What makes something a complex system is that the properties of the system as a whole differ vastly from what appears at first glance to be the sum of their parts. Take starling flocks, for example: an individual starling maintains eye contacts with a few neighbours to determine its flight path, yet collectively, hundreds and thousands of starlings give rise to mesmerizingly beautiful flying patterns of the whole flock in the dusk sky.”

Panja’s research team will investigate ways to minimise the effects of disturbances on the complex system that is the Dutch railway network. “Compare the railway network with stock markets, another classic example of a complex system,” Panja explains. “People are continuously bidding, selling and buying. What happens when a computer glitch causes automatic selling and one cannot stop it? Such an example raises the overarching question: if something unexpected like that happens, how robust is the system? Is it going to crash, or are there ways to limit the damage? This defines the concept of resilience, and in this project we are going to look for improving the resilience of the Dutch railway network.”