NAME Hybrid modeling for the optimization of manufacturing processes

The optimization of manufacturing processes often requires creating various types of classical models (analytical, semi-analytical, FEM, etc.) to replicate the characteristics of these processes. As industry in the Basque Country demands ever-increasing productivity, manufacturing processes grow more complex, and developing these models requires more advanced simulation tools. These models help in understanding the underlying physics and serve as a starting point for designing optimized production strategies (machining, inspection, measurement, etc.). However, industrial implementation is frequently hindered by factors such as high computational cost or reliance on numerous parameters. Meanwhile, the application of Artificial Intelligence in manufacturing processes is becoming increasingly important: AI models can be deployed directly on the shop floor with minimal computational cost and can dynamically adapt to changing manufacturing conditions, which greatly expands their industrial applicability. Nevertheless, a key challenge for the industry is the lack of high-quality experimental data, which limits the accuracy and reliability of AI models. To address these challenges, this asset is defined to bring together both modeling approaches (classical models and AI), fostering the development of customized simulation tools that support the improvement of manufacturing processes (e.g., machining, inspection, measurement) and facilitate their industrial implementation. In this regard, various strategies are considered, such as creating surrogate models based on physics-based models, generating synthetic data to enhance the quality of AI models, and developing generative AI models capable of adapting to new conditions. Hence, this asset provides justification and support for an integrated vision: from classical simulation to synthetic data generation and the development of AI models, all with the goal of optimizing manufacturing processes and accelerating their transfer to industry.