In recent years, accurate 6-DOF (six degrees of freedom) pose estimation has emerged as a pivotal technology in manufacturing, enabling the precise localization and manipulation of objects in complex environments. The effectiveness of 6-DOF pose estimation algorithms critically depends on the availability of diverse, well-annotated datasets. However, obtaining and annotating such datasets present significant challenges due to their scarcity and the intensive labor required for accurate labeling. To address these issues, we propose an innovative approach that employs synthetic data generation, powered by generative artificial intelligence (AI) techniques specifically tailored for industrial applications. Our method enhances the synthetic data generation process by utilizing generative adversarial networks (GANs), which infuse the data with contextual details relevant to manufacturing environments. This process is further augmented by advanced rendering techniques and simulations that create realistic industrial scenes, complete with accurately annotated ground truth for 6-DOF poses. We validate the effectiveness and robustness of our proposed solution through its application in a real-world industrial use case, demonstrating its potential to substantially improve 6-DOF pose estimation in a manufacturing case, used for robotic picking of electronic parts.