Simulation and Modeling of a Data Twin Service for the Autoclave Curing Process

Abstract

The production of composite material components is a high priority in the aerospace and defense industry. It is important to introduce Industry 4.0 related technologies into this field for the purpose of innovating the factory system. Among the processes of the composite materials production, autoclave curing is the key to yielding durable and sustainable structures. However, the settings of the autoclave curing process still heavily rely on the experienced operators in the current practices. In this paper, we adopt the concept of Data Twin Service to allow the interaction between human intelligence and artificial intelligence. The developed Data Twin Service for the autoclave curing process enables the operators to simulate the placement of molds and learns to predict the curing time. Our service is a GUI application which employs a human-in-the-loop design approach and helps users perceive the simulation process intuitively. We propose a two-stage machine learning model to learn the curing time according to the parts and placement. The empirical study is conducted on the one-year historical data and has proven the practical feasibility of the proposed approach. Moreover, the service is currently being used in the manufacturing site and obtains satisfactory performance.

Keywords

• Industry 40
• data twin service
• autoclave curing
• human intelligence
• artificial intelligence
• machine learning