In this research, we investigate an advanced methodology for process monitoring in manufacturing that employs synthetic data with deep learning to enhance production and process quality monitoring. Traditionally, process supervision in the manufacturing industry has depended on manual checks and physical sensors, which are costly, time-consuming, and prone to errors. The study introduces a new approach that utilizes synthetic data to train deep learning models for precise defect detection in 3D-printed objects. We further recommend diverse data augmentation and fine-tuning techniques to improve model efficacy. Our evaluation demonstrates an impressive 90% overall accuracy in real-object defect detection, with a 93% precision rate in identifying defects. These results suggest that models trained on synthetic data are effectively equivalent to those trained on real-world data, indicating that the technique used to generate synthetic data is pivotal to model performance. This innovative strategy mitigates the challenges associated with conventional process monitoring methods, and addresses the practical challenges of applying such technological advancements in actual manufacturing environments. By enabling the creation of extensive training datasets and applying deep learning algorithms for object and pattern recognition, this approach boosts the precision and efficiency of operational supervision within the manufacturing process. The approach and findings of this study significantly enhances efficiency, accuracy, and the overall effectiveness of quality assurance in manufacturing.