This study evaluates a Retrieval-Augmented Generation (RAG) system that combines LLMs with multi-format data to revolutionize industrial maintenance support.

• Tested 8 different LLMs for maintenance assistance applications
• Advanced models like GPT-4 and GPT-4o-mini demonstrated superior performance
• System processes diverse data formats to generate context-aware maintenance instructions
• Enables real-time, accurate support for engineering personnel in complex environments

This research offers significant value for engineering operations by providing more efficient maintenance workflows, reducing downtime, and improving technician effectiveness through contextually relevant assistance.

Cross-Format Retrieval-Augmented Generation in XR with LLMs for Context-Aware Maintenance Assistance