Core technology architecture

1. Whole process data awareness

The mold is built-in 32 distributed fiber optic sensors, which can collect in real time:

• Temperature field distribution (accuracy of ±0.5℃)
• Resin curing pressure (accuracy of ±0.02MPa)
• mold deformation data (accuracy of ±0.02MPa)
• Temperature field distribution (accuracy of ±0.5℃)
• Mold deformation data (accuracy of ±0.02MPa)

• Mold deformation data (accuracy ±0.05mm)

2. Digital Twin Modeling

Built based on ANSYS Twin Builder to build a digital twin of the mold, and through 2000+ hours of thermal-force coupling simulation training, to achieve:

• Predict the material flow behavior of molding processes. Behavior
• Simulate the impact of different process parameters on accuracy
• Real-time mapping of physical mold state changes

3. AI Closed Loop Control

Using Deep Reinforcement Learning algorithms to dynamically optimize:

• Vacuum Pressure Profile
• Heating Rate Control
• Demolding timing determination

Test data shows that the AI algorithm improves dimensional consistency by 67%.

Manufacturing process innovation