Speaker
Description
The efficient operation of accelerator facilities increasingly relies on rapid access to heterogeneous operational knowledge, including logbooks, interlock reports, machine parameters, and historical archive data. At ELBE (Electron Linac for beams with high Brilliance and low Emittance) we are developing a Retrieval-Augmented Generation (RAG) framework that integrates facility documentation and operational records into a unified AI-assisted support tool for operators. The system indexes the electronic logbook, machine archive data, and subsystem manuals using domain-specific embeddings and a vector database. User queries are processed through a large language model that retrieves the most relevant operational context and produces structured, operator-oriented responses. Early tests show that the RAG system can identify relevant past machine states, extract temporal correlations from archive data, and summarize fault patterns that are typically time-consuming for operators to investigate manually. This contribution presents the system architecture, and data integration challenges toward real-time assistance for accelerator operation.
| In which format do you inted to submit your paper? | LaTeX |
|---|