Concept for large-scale evaluation of spatial relations for road infrastructure management: a graph-based federated digital twin approach for comprehensive querying of heterogeneous infrastructure datasets

The maintenance of road infrastructure is particularly challenging due to the various heterogeneous subsystems usually considered in an isolated fashion. This separate examination hinders factoring in effects arising from the interaction of the subsystems. To enable analysis of mutual influences between elements of different infrastructure assets, this paper presents a graph-based approach aiming for scalable evaluation of spatial relations. Focusing on relations between elements associated with differently directed assets, a mapping of spatial references between different linear reference systems is proposed. This mapping approach allows for the analysis of qualitative spatial relations between components of different assets, like roads and bridges, particularly in overpass situations, thus enabling consideration of mutually influencing effects between them. For validation, we use real-world data from German infrastructure management and showcase how winter maintenance data can be associated with bridge damage documentation.