BIM‐Based Information Architecture for Object‐Centered Normalization and Time‐Series Structuring of Bridge Maintenance Information

The adoption of digital platforms for managing bridge inspection and maintenance histories has increased in recent years. However, many conventional digital shadow‐based systems remain focused on the unidirectional recording or visualization of inspection results. In such systems, maintenance information is often stored in document‐, table‐, or image‐based formats, and object relationships and temporal histories are not explicitly structured. These limitations make it difficult to accumulate and compare component‐level maintenance records over time and to reuse them as structured inputs for subsequent analytical applications. To address these limitations, this study proposes a building information modeling (BIM)‐based information architecture for object‐centered normalization and time‐series structuring of bridge maintenance information. The proposed architecture consists of three layers: a model layer for object ID‐based data normalization and temporal encoding, an analytics‐compatible layer for separating data transformation from future analytical modules, and a service layer for spatial visualization and maintenance history retrieval. The architecture was implemented as a web‐based platform and applied to real inspection and maintenance records through JSON‐based data linkage. The implementation examined the operational feasibility of object ID‐based mapping, cumulative time‐series management, and three‐dimensional visualization. Rather than developing a complete digital twin (DT) or validating a predictive AI model, this study provides an intermediate information architecture that supports data‐structural readiness for future analytical extension by organizing maintenance records around objects, attributes, and time.