Gerta Köster researches dynamic systems that form where many people come together. The goal of her "Crowd Dynamics" research group is to use computers to make daily life safer and more efficient. To this end, the team models, observes phenomena such as traffic jams and recreates them in a virtual world. In parallel, the researchers have an operator-based, explainable AI learn substitute models. Current research questions are: How many passengers need to follow navigation instructions to equalize crowds at the Münchner Freiheit subway station? And, how is exposure to Sars Cov-2 viruses at ticket counters related to queue management?