Manufacturing defect mitigation focuses on identifying the causes of defects and reducing their occurrence by adjusting process configurations. This requires the development of a smart decision-making system that can assist in self-configuration and self-optimization of a production line. To achieve this, this paper proposes a case-based domain knowledge modeling and reasoning approach for cognitive intelligent defect mitigation using knowledge graphs and generative artificial intelligence (GenAI). Unlike conventional case-based reasoning, the proposed approach employes knowledge graph database for knowledge modeling and case representation, enhancing the efficiency of case retrieval. Additionally, the large-language model (LLM) is employed to assist in knowledge acquisition and knowledge reasoning for case adaptation. An example of defect mitigation in an assembly line with in-process inspection is presented to validate the feasibility and effectiveness of the proposed approach.
GPT-Powered Case-based Domain Knowledge Modeling and Reasoning for Cognitive Intelligent Manufacturing Defect Mitigation
5 Views
5 Downloads