Recommending engineering knowledge with context-aware agentic AI: a case study in on-site wind turbines assembly

With the rapid development of the renewable energy industry, the demand for on-site assembly of wind turbines has been increasing. However, the assembly process of wind turbines involves complex process knowledge and high contextual constraints, and traditional knowledge recommendation methods have certain limitations in addressing these challenges. To improve the efficiency and accuracy of workers in acquiring specialised knowledge at the assembly site, this paper proposes a context-driven multi-agent collaborative solution generation method. This method considers contextual constraints and interdisciplinary collaboration issues by constructing a multi-agent system that combines specialised knowledge of wind turbine assembly with situational information, thereby recommending task-specific knowledge. We constructed a context-oriented knowledge graph for wind turbine assembly, structurally modelling key processes, procedures, steps, and situational constraints. We propose a multi-strategy knowledge recommendation mechanism, which selects appropriate query strategy based on different task scenarios. We introduce expertise degree and agent memory mechanism, adjusting the agent's memory structure and decision-making approach to ensure the accuracy and consistency of information transmission. Experimental results show that the proposed method demonstrates high accuracy and effectiveness in knowledge recommendation for wind turbine assembly, providing intelligent and customised solutions for assembly personnel.