Currently, Sebastian is trying to characterise the structural patterns that can emerge in urban mobility data across multiple scales. 

Understanding human mobility patterns has long attracted interest in several areas of urban planning and design, as it can help create more efficient, sustainable, and equitable cities by guiding decision-making processes, optimising transportation systems, and fostering a better understanding of how people move within urban areas. However, current methodologies used to infer these patterns typically exhibit several limitations, such as difficulties integrating different data sources, the inability to operate at multiple scales, and complications when attempting to incorporate contextual information. On the other hand, most principled statistical inference methods that allow to overcome these limitations can prove ineffective when applied to spatial interaction data. The aims of the project are then twofold: to assess the efficacy of currently existing models in inferring urban mobility patterns and to develop novel generative models that can incorporate spatial information in a principled way.