Topology-Driven Semantic Integration of Independent BIM Models for Renovation Analysis

This paper proposes a topology-driven semantic integration approach to align as-is and to-be BIM models in building renovation scenarios. Starting from independent IFC files, a surface-based topological analysis infers physically meaningful contacts between envelope objects using coplanarity, opposite orientation, vertical overlap, and an overlap-coverage metric. The detected adjacencies are then represented as cross-model semantic relations by instantiating bot:Interface links between the corresponding bot:Element individuals in a dedicated knowledge graph, preserving model independence. The resulting topology enables cross-model SPARQL queries to extract and aggregate energy-related data at the envelopesystem level without IFC merging or manual data preparation.