Towards flexible and lightweight solutions for Urban Digital Twins

Urban Digital Twins (UDTs) are increasingly used to support analysis, planning and decision-making in complex urban environments. However, many existing solutions rely on rigid and monolithic architectures that limit their capacity to evolve with changing data sources, applications and technologies. This paper presents the Urban Digital Twin research activities conducted within the GeoScITY laboratory, focusing on architectural principles and tools that promote flexibility, accessibility and long-term maintainability. These contributions rely on semantic 3D city models as a stable spatial backbone and combine lightweight encodings, semantic validation middleware and automated workflows to integrate static and dynamic urban data. The approach is illustrated through several tools developed at GeoScITY, including City2Twin for managing and visualizing evolving 3D city models, automated workflows for data integration, CERBERE for schema mediation and semantic validation, and GIS4IoRT for integrating Internet of Things and robotic systems into geospatial digital twins. Together, these contributions demonstrate how Urban Digital Twins can be designed as evolving, interoperable and reusable systems rather than fixed digital replicas.