Posters
Come meet the poster presenters to ask them questions and discuss their work
We would like to invite you to come and see the posters at our upcoming conference. The posters will showcase a diverse range of research topics, and will give delegates an opportunity to engage with the authors and learn more about their work. Whether you are a seasoned researcher or simply curious about the latest developments in your field, we believe that the posters will offer something of interest to everyone. So please join us at the conference and take advantage of this opportunity to learn and engage with your peers in industry and the academic community.
PO013: Multi-Agentic Collective Intelligence with Generative AI for Interactive Decision Support in the Offshore Wind Industry
Idris Olawale Alugo, PhD Researcher, Warsaw University of Technology | Senior Data Scientist, Warsaw University of Technology
Abstract
As offshore wind energy continues to proliferate in popularity globally with wider data abundance and complexity, there are growing challenges associated with Operations and Maintenance (O&M). Many studies have demonstrated the potential of machine learning (ML) and deep learning models in predicting incipient faults and power output from Supervisory Control and Data Acquisition (SCADA) and sensor readings. However, despite their accuracy, these models are generally trained from isolated datasets that fail to capture nuanced inter-dependencies between e.g. power output and blade health, maintenance planning and historical work orders, etc. Generative AI (GenAI) models and Agentic AI frameworks have demonstrated immense potential in domains like finance and healthcare recently, owing to the ability of Large Language Models (LLMs) to harness collective intelligence with a human-in-the-loop. However, whilst powerful, LLMs are prone to hallucinations, which limit their practical use in safety-critical engineering systems like wind turbines. We present a GenAI prototype framework for collective intelligence in the wind industry combining predictive insights from four specialised agents on fault prediction, power forecasting, offshore vessel transfer and structural health monitoring. Outputs from these components are collectively fed as context to a domain-specific LLM, which allows human engineers to pursue interactive question-answering in natural language, querying turbine status, plan O&M and vessel transfers, access historical work orders, amongst others. We achieve reliability by integrating our LLM with a domain-specific knowledge graph (KG), XAI4Wind, which captures and retrieves expert knowledge and offers reasoning trails for each query to increase transparency and trustworthiness. Our model can generate SAP-style work orders to facilitate O&M planning and inventory management. We envision that this framework offers a step forward in addressing concerns with leveraging natively non-deterministic GenAI models with confidence in the wind industry, supporting interactive and explainable human-AI collaborative decision making for managing offshore wind assets.
No recording available for this poster.
