Digital twin model-based smart assembly strategy design and precision evaluation for PCB kit-box build
Xurui Li,Guangshuai Liu,Si Sun,Wenyu Yi,Bailin Li
DOI: https://doi.org/10.1016/j.jmsy.2023.09.010
IF: 12.1
2023-12-01
Journal of Manufacturing Systems
Abstract:Research concerning microelectronics assembly has attracted increasing attention from the manufacturing industry. In this context, achieving precise placement of the printed circuit board (PCB) inside the enclosure is the most critical aspect of the kit-box build assembly task. However, automating the PCB kit-box build assembly (PKBA) process remains a challenging work. This study presents a Digital Twin (DT) model that enables accurate perception of object pose for intelligent PKBA assembly, facilitating real-time monitoring and evaluation of its service status. In our system, a symmetry-drive method is proposed to optimize the initial pose in 6-DoF matching during DT assembly. Based on the developed technology, a three-stage learning method is established to achieve grasping point localization and robotic assembly trajectory planning. To ensure accurate and robust robot assembly, we developed a PKBA quality prediction model based on the DT system, which predicts the uncertainty of actual PCB assembly positioning by small displacement torsor (SDT) theory and Monte Carlo methods. Particularly, the assembly quality of the actual product is effectively monitored when the state of the virtual simulation model corresponds to the physical assembly object. Finally, a prototype system and a case study involving dexterous assembly tasks are conducted to verify the effectiveness and feasibility of the proposed method. The results indicate that the proposed PKBA strategy achieves an 82% assembly success rate. By employing well-designed strategies, our method ensures that the majority of errors are below 0.8 mm and 0.6 degrees.
engineering, manufacturing, industrial,operations research & management science