SS07: LLMs and Machine Learning for Industry 4.0. A Use Case in Digital Transformation
Francisco de Arriba Pérez
University of Vigo, Spain
Abstract
Industry 4.0, combined with Machine Learning, sensorization, and new generative AI tools, is revolutionizing process automation and optimization, giving way to Industry 5.0. The inclusion of language models (LLMs) has transformed human-machine interaction in highly sensorized environments. These models improve decision-making, anomaly detection, predictive maintenance, and process optimization. These models facilitate the automation of technical documentation, reporting, and real-time assistance for workers within the industrial ecosystem.
Objectives:
- Optimized LLMs for worker Key Performance Indicators (KPIs) detection
- Exploring how LLMs can analyze KPIs
- LLMs in the context of industry 4.0: relevant use cases
- LLMs for the digital transformation of industrial processes, enhancing automation and decision-making
- Predictive maintenance, quality control, documentation automation, and operator assistance using LLMs
- LLMs for predictive maintenance, ensuring product quality, and assisting human operators in industrial environments
- Challenges and opportunities for LLMs in security, privacy, and adaptation to specific industrial environments
- Risks and benefits of LLMs, particularly regarding data security, privacy concerns, and model customization for industry-specific needs
Organizer
Dr. Francisco de Arriba Pérez‘s primary research focuses on integrating Machine Learning (ML) and Natural Language Processing (NLP) techniques. He is the author of 30 open-access publications in JCR and SJR-indexed journals. His research work has accumulated 600 citations, with an h-index of 12. Additionally, he has participated in 22 conferences, is a member of 4 international scientific committees, and is a guest editor for 3 special issues. In recent years, he has expanded his work into applying LLMs in multidisciplinary environments. Specifically, he has participated in 21 competitively funded projects, including 7 European projects, and has successfully managed over 24 industry-collaborative projects across the legal, financial, and industrial sectors. Currently, he serves as the principal investigator for the PERTE project (€95,000.00), focusing on predictive maintenance in the agriculture sector. Recently, he has explored the use of LLMs and ML for worker KPI detection in industrial environments, failure detection, and predictive maintenance.