Optimizing Aerospace Product Maintenance A Novel Multi-Modal Knowledge Graph and LLM Approach for Enhanced Decision Support
Conference paper
Authors | Awill, R., Khan, W., Hussain, M. and Anderson, B. |
---|---|
Type | Conference paper |
Abstract | Siloed and inaccessible repair knowledge hinders the efficient maintenance of critical Turbine Engine components in the aerospace industry. This research introduces a novel multi-modal knowledge graph, leveraging Natural Language Processing (NLP) and Large Language Models (LLMs) to extract and structure repair rules from unstructured documents into a 131-node, 148-relationship graph. This advancement enables immediate access to essential information and facilitates datadriven decision-making, enhancing repair accuracy and efficiency. Implemented at AddQual Ltd., the knowledge graph reduced information retrieval times by 70%, increased repair speed by 20%, and is projected to yield 20% annual cost savings. These results highlight the transformative potential of integrating AI with knowledge graphs in aerospace maintenance. Future work will focus on advancing robust data validation frameworks and developing adaptive AI algorithms, extending the benefits across the aerospace sector and beyond. |
Keywords | Knowledge Graph; Multi-Modal; Turbine Repair; Decision Support; Aerospace; Text2Cypher |
Year | 2024 |
Conference | The Extended Semantic Web Conference 2024: Fabrics of Knowledge: Knowledge Graphs and Generative AI |
Publisher | The Extended Semantic Web |
Web address (URL) | https://2024.eswc-conferences.org/wp-content/uploads/2024/05/77770127.pdf |
Output status | Published |
Publication dates | 29 May 2024 |
Publication process dates | |
Deposited | 27 Jun 2024 |
https://repository.derby.ac.uk/item/q70z5/optimizing-aerospace-product-maintenance-a-novel-multi-modal-knowledge-graph-and-llm-approach-for-enhanced-decision-support
114
total views0
total downloads3
views this month0
downloads this month