JuPedSim @ SciCoCo.nrw 2026

My colleague Pit Steinbach and I are going to speak about JuPedSim at SciCoCo.nrw 2026!

Outcomes We Can Sell: The JuPedSim Story

We want to tell you a story about JuPedSim, a pedestrian dynamics simulator. It is a story about success, failure and renewal. How JuPedSim grew out of several PhD projects, almost collapsed under its own weight, and was saved by a team that committed to its long-term future and dared to try something different. JuPedSim supported the ideas behind an ERC Synergy Grant as they took shape, and remains part of that science today.[1][2]

If you attend this conference, it is very likely you believe “Better Software, Better Research.” But our personal convictions alone do not unlock budgets. We are convinced our community needs better stories when engaging stakeholders. Stories that exemplify that software done well is an enabler. Stories that help leadership connect the dots and see the long-term impact better software has.

By 2020 JuPedSim had grown into a collection of hard-to-maintain executables driven by complex XML files: powerful, but opaque, and increasingly inaccessible to new users and new developers. Without ever intending to compete with commercial vendors, we had drifted into building everything ourselves, even our own CAD-like editor. But a complete end-to-end solution is what a vendor must build; a research group does not have to. We needed to rethink what JuPedSim was for and who it was for. The answer was a strategic one: JuPedSim would focus on its core mission of model development and fundamental research and leave everything else to tools that already do it well. Everything that followed was a consequence of this decision. Features that did not serve the core were cut without replacement. What remained became a focused Python package: fewer things, done better. Because behaviour was now expressed in code rather than fixed in configuration files, parameters could change while a simulation was running. This made it possible to couple pedestrian dynamics with models of social behaviour, letting agents adapt their movement to changing social conditions throughout a simulation. These capabilities opened up new scientific questions and helped shape the scientific ideas that later became part of an ERC Synergy Grant.

We want to tell you our success story, and we want you to look for your own success stories that can be told to stakeholders. We want to encourage you to evaluate your Research Software Engineering work through the lens of “Do I add value for my stakeholders?” Do this with the aim of speaking their language, so as to open not just their hearts and minds but also their purses. We need to find and convincingly tell those success stories to drive the point home that “Better Software enables Better Research” and that groups who understand this will advance faster and leave behind those who still see no value in RSE.

References
[1] https://JuPedSim.org
[2] https://www.fz-juelich.de/en/news/archive/press-release/2025/understanding-crowds-erc-synergy-grant-awarded-to-research-team-from-julich-and-groningen