Real-Time Data Analysis with Artificial Intelligence in Parts Manufactured by FDM Printer Using Image Processing Method

Abstract

Abstract In this study, samples manufactured with polylactic acid (PLA) plastic material using the fused deposition modeling (FDM) type printer were analyzed during the manufacturing process using image processing and real-time big data analysis. The purpose of real-time big data analysis is to provide an effective and efficient guide to the user in the manufacturing process regarding the manufactured part’s mechanical properties. In this study, compression samples were prepared according to ASTM D695-15, Standard Test Method for Compressive Properties of Rigid Plastics, test standards and subjected to mechanical tests. In the first stage of the research, using artificial neural networks (ANNs), processing parameters were estimated with 92.5 % accuracy according to the R2 performance evaluation criterion. In the second stage, each layer’s infill percentage and layer thickness of the compression sample were analyzed using image processing techniques. In the final stage of the study, using the Python programming language, a user-specific visual interface is designed for showing the results and graphics related to the material processing step in FDM 3D printing.