Transfer of implicit semi-formal textual location descriptions in three-dimensional model contexts

Abstract: The current bridge inspection documentation in Germany is only text-based. Thus, the locations of damages and components are defined in implicit, semi-formal textual descriptions, which depend on the viewing direction of the inspecting engineer. This paper presents an approach to automatically transfer these location descriptions into explicit three-dimensional model elements and conceptual damage representations. The approach works with rule-based transformations encoded in dedicated algorithms and achieves unambiguous model-based localization and geometrical representation of the documented inspection data. Concurrently, the process links legacy information with new model data, contributing to an efficient implementation of digital twins.