Optimization of Production Infrastructure processes, using Artificial Intelligence (AI) Methods (OPTIMUM)
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This study presents the OPTIMUM framework, a machine learning-driven system, for optimizing control units in manufacturing environments.By leveraging real production data, we implemented and evaluated four key models: delay prediction, anomaly detection, efficiency optimization and prescriptive decision-making.A Random Forest classifier predicted production delays using early-stage features, while an Isolation Forest model identified operational anomalies without labeled data.A regression model accurately predicted air consumption with an R² of 0.9991, supporting proactive resource management.Finally, a decision tree model generated interpretable rules for minimizing delays through scheduling adjustments.The results demonstrate that machine learning techniques can not only detect inefficiencies but also recommend actionable adjustments, significantly improving responsiveness and sustainability in production systems.The proposed framework offers a modular and scalable approach to integrating AI into control units, paving the way for smarter, data-driven manufacturing operations.
Fields of Study & Subjects
Impact of AI and Big Data on Business and Society
Digital Transformation in Industry
Management and Optimization Techniques
Publication details
- Journal / source
- Acta Polytechnica Hungarica
- Volume
- 23
- Issue
- 3
- Pages
- 99–120
- Publication date
- 1 January 2026
- Language
- EN