Digital Twin Architecture to Use for Optimizing an AoD-Printing Process

Digital Twins have been gaining relevancy over the past decade as computing power becomes increasingly more powerful and cheaper. However, most digital twins require an already profound knowledge of their physical counterpart in order to mimic it. This paper proposes a data architecture which enables a digital twin to learn from its physical counterpart and help to build models in applications where little established models are available. The digital twin in this paper focuses on accompanying an additive manufacturing process.