AI Optimization of Resin Infusion in Composite Manufacturing

Use Case:

AI-assisted optimization of catalyst and accelerator quantities for vacuum infusion processes in composite manufacturing reducing defects improving energy efficiency and supporting adaptability across products.

Outcome:

• Outstanding accuracy results;
• Generalizability of the AI model validated on different part sizes;
• Reduced waste and optimized resin use;
• Fully automated resin mix computation with real-time operator dashboard;
• Continuous improvement pipeline via feature extraction and Bayesian inference.

Ecosystem Support:

• StairwAI project support;
• Local AI experts for DevOps pipeline and sensor integration;
• Access to advanced sensing technologies (e.g., OPTIMOLD).

AI Relevance:

This success story exemplifies AI democratization for SMEs through:

• Low-barrier AI deployment via dashboards and APIs;
• Adaptable models trained with small datasets and expert annotation;
• Interoperability of AI and physical systems (sensors, cameras);
• Use of AI to enhance manufacturing robustness, efficiency, and sustainability.

Summary:

Trygons SA, a Greek SME specialized in composite parts production, tackled the challenge of optimizing the vacuum infusion process—a crucial and delicate step in manufacturing composite components. Through collaboration within the StairwAI project, Trygons developed AgRerTry, a dual-AI model system powered by advanced sensor integration and real-time video streams from RGBT cameras. This AI-driven platform automatically determines the ideal catalyst and accelerator quantities based on ambient temperature and resin volume, streamlining the infusion setup. An intuitive dashboard enables operators to monitor and interact with the process, reducing material waste and human error. After successful validation, the AI system proved generalizable across different product types. The system significantly reduced failed infusions (0% in October) and laid the foundation for broader adoption in resin-based manufacturing. This innovation was enabled by local AI experts and support infrastructure provided through the StairwAI ecosystem, which guided sensor integration, DevOps deployment, and dashboard development.