A Conceptual Framework for Updating Urban Infrastructure in City Digital Twins

Abstract. This work presents a conceptual framework for updating urban infrastructure by separating the update process into two loops, the data loop and the application loop. The key aspect of the data loop is intelligent change detection followed by validation strategies to maintain an up-to-date catalogue of city infrastructure. The updated catalogue can subsequently utilised to create user-centric digital twins. The advantages of using such a framework are presented, in addition to discussing the challenges and their possible solutions. Further, the scope of integrating advanced change detection approaches, new data structures such as Knowledge Graphs and Voxels, and AI methods such as the recent Large Language Models (LLM) and Continual Learning are discussed.