IIC Journal of Innovation 12th Edition | Page 32

Digital Twin Development for Serial Manipulators: Data Driven Optimized Planning and Sequencing of Tasks background studies, such as developing frameworks for applications of DT. 10 11 planning and simulation are executed to teach the robot with expected movements. Subsequently, the virtual robot setup enables engineers to review and verify the new robot motion path. Finally, the approved path is released to the robot controller linked with the programmable logic controller (PLC) of the assembly line to execute assembly operations. During training and operation, the cobot keeps publishing its critical status information to the virtual environment in real time. By following these steps, which are also aligned with the vision of a cyber-physical system implementation in Industry 4.0, 13 14 the order is executed with efficient monitoring, planning, simulation and optimization. Such an implementation is novel as it is generic and can be used for any serial manipulators. This differs from the state of the art which is targeted towards specific machine or work cell DT POCs. Although studies on DT are various, its prevalent implementation has not been realized. One of the main reasons is that companies still encounter challenges to identify the implementation scope where DT will create worthwhile business value. 12 Motivated by the above research gap, in this article, a practical use case of DT developed and implemented in the Model Factory (MF) program at Advanced Remanufacturing and Technology Centre (ARTC), Singapore, is introduced. In this use case, a DT is developed for a gearbox assembly line automated by a collaborative robot (cobot) for new gearbox sub-assembly configuration requirements. Once the customer order containing customized configuration of gearbox sub-assembly is received, new assembly and component designs are trained to build with an object recognition model using machine vision and deep learning algorithms. The trained model is integrated with the desired operation sequence to pass the information to the robot motion planning software. Next, robot In the first section of this article, the definition and advantages of a DT model will be recapped. Next, the DT model developed by ARTC will be presented. In particular, the novelty and innovation of the development of the DT model will be discussed in detail. 10 Söderberg, R., Wärmefjord, K., Madrid, J., Lorin, S., Forslund, A., & Lindkvist, L. (2018). An Information and Simulation Framework for Increased Quality in Welded Components, CIRP Annals, 67:1, 165–168. 11 Zhuang, C., Liu, J., & Xiong, H. (2018). Digital twin-based smart production management and control framework for the complex product assembly shop-floor. The International Journal of Advanced Manufacturing Technology, 96(1-4), 1149-1163. 12 Parrott, A., & Warshaw, L. (2017), Industry 4.0 and the digital twin. Deloitte University Press. 13 Luo, W., Hu, T., Zhang, C., & Wei, Y. (2019). Digital twin for CNC machine tool: modeling and using strategy. Journal of Ambient Intelligence and Humanized Computing, 10(3), 1129-1140. 14 Zhao, R., Yan, D., Liu, Q., Leng, J., Wan, J., Chen, X., & Zhang, X. (2019). Digital Twin-Driven Cyber-Physical System for Autonomously Controlling of Micro Punching System. IEEE Access, 7, 9459-9469. IIC Journal of Innovation - 27 -