Defect trigger identification in a construction digital twin using quality linked data and iterative deviation computation

Current quality management in construction faces several challenges, including insufficient precision in control, unclear allocation of quality responsibilities, and a lack of long-term management mechanisms. As the industry shifts from a problem-solving after occurrence approach to a proactive paradigm, identification of potential quality defects and risks has become a primary prerequisite. This study introduces the concept of quality defect triggers, detectable, modelable, and controllable precursors to quality issues that operationalize risk factors and can be visualized within a construction digital twin (DT). Building on an exploration of their evolutionary mechanisms, the study develops an integrated framework that combines ontology and linked data technologies to identify quality defect triggers through multisource heterogeneous data integration, comprising three core components: (1) the evolutionary mechanisms of defect trigger as the theoretical foundation and underlying principles for computation; (2) an ontology-based four-layer semantic model encompassing quality standards, quality data, defect triggers, and construction processes, implemented using the GraphDB graph database to achieve semantic linkage; and (3) a quality defect trigger identification and iterative deviation computation engine in construction DT. By unifying data modeling, semantic mapping, and automated reasoning, the framework provides a systematic and practical approach to early quality defect identification. Its effectiveness was validated through a real-world prefabricated composite-slab installation case, comparative experiments, and expert evaluation, demonstrating its applicability and potential to support proactive, data-informed quality management in construction practice.