Enabling Spatial and Temporal Correlation of Heterogeneous Infrastructure Data Using Knowledge Graphs

Abstract: Road infrastructure is a complex system with interconnected subsystems, making the assessment of condition correlation crucial for predictive maintenance. However, the substantial heterogeneity of these subsystems complicates the correlation of existing data sets. To address this challenge, we propose a conceptual framework representing data sets and their spatial and temporal relationships through graphs by enhancing existing approaches with considerations of temporal dependencies. Hence, large-scale correlation analyses between heterogeneous subsystems and their related data are enabled. Validation is performed using real-world data from German infrastructure management, specifically associating traffic count data with bridge conditions based on their spatial and temporal relations.