Generative and Predictive AI for digital twin systems in manufacturing.

Area of Science:

• Manufacturing Technology
• Artificial Intelligence
• Digital Twin Technology

Background:

• Modern manufacturing increasingly relies on AI and Digital Twin (DT) integration for enhanced operational capabilities.
• Real-time monitoring, predictive maintenance, and process optimization are key drivers for smart manufacturing advancements.

Purpose of the Study:

• To design and partially implement an AI-enabled Digital Twin (AI-DT) system for manufacturing applications.
• To leverage Generative AI (GAI) and Predictive AI (PAI) for specific manufacturing tasks within the AI-DT framework.

Main Methods:

• Deployment of Generative AI (GAI) for data augmentation, geometric inspection, and creating virtual testing environments.
• Utilization of Predictive AI (PAI) with sensor data for proactive defect detection and quality analysis in welding.
• Integration of GAI and PAI modules within a Digital Twin framework for manufacturing.

Main Results:

• The AI-DT system demonstrates capabilities in augmenting training data and generating 3D virtual testing environments.
• PAI effectively enables proactive defect detection and predictive quality analysis in welding processes.
• The integrated system enhances issue anticipation and supports manufacturing decision-making.

Conclusions:

• The developed AI-DT system provides a robust foundation for scalable, intelligent digital twins in smart manufacturing.
• The integration of AI modules significantly improves operational efficiency and quality assurance.
• This work advances early-stage digital-physical convergence in manufacturing settings.