Xizhong Zhang
Information Management Department, China Construction Science & Technology Group CO., LTD Beijing, China,100070.
Xinliang Duan
Information Management Department, China Construction Science & Technology Group CO., LTD Beijing, China,100070.

Abstract:

Prefabricated concrete buildings are gaining widespread adoption due to their high efficiency and sustainability. However, managing the vast and heterogeneous quality data generated throughout the component lifecycle—from design to maintenance—poses significant challenges. This paper proposes an intelligent data transformation and application framework based on data lineage modeling and large-scale artificial intelligence (AI) models. The framework incorporates semantic parsing, multi-source data integration, and standardized data modeling through metadata and knowledge graph technology. A data root-node-driven metamodel is constructed to trace and associate design, production, and assembly data with each component. Furthermore, a deep learning-based large model is introduced to automate the transformation of multi-format quality data across systems. A case study from a large industrial park project demonstrates the framework’s effectiveness in improving data quality traceability, integration efficiency, and intelligent decision support.