Vision-Based Monitoring and Control for 3D Printing Process with Dynamic ROI and Path Modification Algorithm

Additive manufacturing of Fused Deposition Modeling (FDM) type 3D printing has rapidly developed. In deposition-based layering process, errors in the lower layers are fatal defects, leading to loss of time and materials. In this study, we propose a motion control method that suppresses failures during the modeling process by using visual feedback control with an RGB camera. A camera set up near the nozzle detects the molding line and generates a new path from the coordinates obtained by the image, allowing for real-time path modification. In this paper, we conducted experiments to investigate the effect of the path modification by the calculation algorithm and the effectiveness of changes in the Region of Interest (ROI) for path modification. Experiments using stacked materials showed that the proposed method is effective in suppressing failures during the printing process by printing several layers. The proposed method is significantly different from conventional control methods in that it detects path changes during modeling and flexibly changes the path in real-time. This vision-based compensation system is expected to contribute to the automation and quality improvement of the additive manufacturing process.