A knowledge graph-driven framework of multi-stakeholder synergistic operation and maintenance for complex products: design, implementation and industrial validation

To ensure the optimal-state operation, low maintenance costs and optimized resource allocation, synergistic operation and maintenance (OM) mode of complex products (CPs) is becoming a current research hotspot. However, multiple stakeholders (e.g. user, manufacturer and service provider) involved in the synergistic OM of CPs are facing challenges in integrating heterogeneous data, resolving ambiguities in service requirements and generating explainable maintenance strategies due to inconsistent data semantics and fragmented knowledge divide. Moreover, although the promising application of knowledge graph (KG) to handle large-scale data semantically and resulting in better knowledge sharing, the research simultaneously combing multiple stakeholder synergistic OM and KG is in its infancy. To address these challenges and problems, a KG-driven framework of multi-stakeholder synergistic OM for CPs is developed is this paper. Then, a formal ontology modelling method for consistent representation of multi-stakeholder heterogeneous data and a Naive Bayes-enabled knowledge navigation model for reducing the ambiguities of multi-stakeholder OM information are proposed to provide technical support for effective implementation of the framework. Finally, an industrial application scenario of an electric multiple units (EMU) trailer bogie OM is presented to validate the feasibility and effectiveness of the framework. The result shows that the developed framework achieves 87.8% accuracy in OM knowledge classification and retrieval (e.g., fault cause identification and maintenance strategy recommendation). Therefore, it can be used to effectively facilitate the heterogeneous data using and implement the multi-source knowledge sharing among different stakeholders for generating explainable and actionable synergistic OM strategies for unfamiliar maintenance tasks.