DARe resilient transport simulation scenarios stand out at MATSim User Meeting

In June, I attended the MATSim User Meeting 2025 (MUM25), hosted at TU Munich, Germany. This is the second consecutive year that DARe has been represented at this event, where developers and users meet to discuss and showcase innovations and projects developed using the agent-based modelling MATSim tool. This year, I was joined by Prof James Woodcock from the DARe Cambridge team, and fellow Newcastle University team members Dr Alistair Ford and Daniel Bell, who presented on some of the research projects that DARe is developing. 

The MATSim meeting featured a full programme of scientific research, software demonstrations, and discussions among researchers, users and practitioners from all around the world.  

DARe’s contribution to the event was twofold. Dr Ford presented ‘Disappearing Traffic: Can MATSim Simulate Travel Behaviour During Disruptive Events?’, sharing the results of a work-in-progress MATSim scenario, analysing the impact of the four-year Tyne Bridge restoration work taking place in the Tyne and Wear region (UK) and other associated infrastructure changes on travel demand. The model results were compared to observed traffic volumes to validate the model’s accuracy and identify how the agents adapted their behaviour compared to the real world. 

Then, I presented ‘A MATSim Modelling Framework to Simulate the Impacts of Extreme Rainfall Events in the UK‘, where the methodology developed to simulate the impact of extreme rainfall events to transport demand was applied to the Tyne and Wear region, simulating the 2012 ‘Toon Monsoon’ flash flood event (when more than the average amount of rainfall for June fell in just two hours on 28 June 2012). Comparisons were made between two extreme scenarios; the first showed how the agents behave as on a normal day (i.e., without the possibility of adapting their behaviours to the consequences of the extreme weather event) (i.e., worst case scenario) and, the second,  when they are allowed to learn through a set of iterations, where modifications to their activity plans (e.g., the use of different transport modes, followed routes and starting time) allowed agents to avoid, when possible, flood-affected or highly congested areas (i.e., ideal case scenario). 

Other presentations shown during the MATSim meeting ranged from the use of machine learning techniques to optimise transport policies; the simulation of human behaviours, transport scenarios and mobility policies applied to different cities; to Demand-Responsive Transport simulations; simulation of public transport modes; and charging behaviours, among many others. (A full list of all presentations can be found here). 

At the end of the event, we were thrilled to receive Best Paper Award, recognising the work we developed to simulate the impact of extreme rainfall events on transport demand. This is a great achievement that highlights the research we do at DARe and enhances the importance of simulating these types of scenarios for a more resilient and sustainable transport sector. Notably, only DARe’s presentations were focused on the effects of extreme weather events and transport resilience, highlighting the criticality of research that considers these elements together and the need for Hubs like DARe.  

All in all, this was a great event that enabled the possibility of learning about what other MATSim users are developing and producing, as well as networking with both new and old friends. Personally, it was a great experience to come back to the city where I spent six months during an Erasmus exchange a few years ago. Danke, München. 

Looking forward to MUM26 in Paris.