SUMMARY

Digital manufacturing seeks to enhance the end-to-end design-to-manufacturing workflow by seamlessly integrating state-of-the-art technologies to enhance efficiency, precision, and agility. However, the progress of digital manufacturing is hindered by extensive human intervention. Recently, machine learning (ML) processes have emerged as a catalyst for advancing digital manufacturing. Through the training of neural networks with curated datasets, numerous intermediate processes previously reliant on human input can now be automated. This study focuses on leveraging machine learning techniques for generative design, emphasizing functionality and manufacturability through the automatic selection of manufacturing processes. The selection procedure for manufacturing processes takes into account shape information and may or may not involve material details. This can be accomplished either through direct classification of processes or by identifying similar designs and extracting manufacturing insights from these similar counterparts. In generative design, two distinct procedures, namely design for functionality (DFF) and design for manufacturability (DFM), emerge. In DFF, the primary goal is to enhance the functionality of the design. During DFM, the functional design serves as input to enhance its manufacturing capabilities. Ultimately, the iterative loop between DFF and DFM contributes to the continuous improvement of design functionality and manufacturing capabilities.