Metal Additive Manufacturing Conference - MAMC 2023

The introduction of additive manufacturing (AM) has dramatically increased design freedom, and at the same time complicated the identification of the most suitable manufacturing technology for mechanical parts. Manufacturing process selection is typically carried out by experts, but this operation is increasingly challenging and experts are difficult to recruit. Hence, the main aim of this research is to differentiate the parts to be fabricated using additive and traditional manufacturing technologies. The proposed methodology uses images of CAD (Computer-aided design) models that are taken from freely accessible web databases. The dataset contains forty images extracted using SOLIDWORKS 2022 software, where 42.50% are AM-ed products. The proposed hierarchical image clustering algorithm had 87.50% accuracy. Thus, this methodology shows great potential for manufacturing process selection and image processing applications.