Ontological reasoning for automated tunnel defect diagnosis and root cause identification

Defect diagnosis and root cause identification are pivotal to safe tunnel operation and the formulation of maintenance strategies. Due to the heterogeneous and multisource nature of defect data, manual integration and analysis remain challenging and inefficient. This paper proposes a framework that includes a Tunnel Defect Analysis Ontology (TDAO) for knowledge extraction from heterogeneous data and a rule-based reasoning procedure that supports automated defect diagnosis and root cause identification. The framework's accuracy and efficiency are demonstrated through a case study on the Shanghai Yangtze River Tunnel, and its practical value is further confirmed by expert interviews. The result shows that the proposed approach offers an automated, economical, and high-efficiency solution that advances intelligent tunnel operation. Future studies can embed maintenance-decision knowledge and additional defect types into the ontology and implement automatic rule generation to enhance the framework's general applicability.