Effectiveness of Digital Twin Framework for Collaborative Robotic Manipulation

This work examines the effectiveness of a digital twin (DT) framework using an industrial pick-and-place case study with a collaborative robotics arm. The problem addressed is the need for improved production process planning and visualization in robotics. The employed method involves creating a DT and evaluating its fidelity to a physical robotic system performing a pick-and-place task. Evaluations included comparing the digital and real robot trajectories, utilizing ISO 9283 performance testing, and analyzing metrics like RMSE, MAPE, and R2. The evaluation shows initial positive results indicating that the proposed DT framework fits the real data well, thus, demonstrates feasibility of this approach. The results can be used to improve the planning and visualization of the production process for collaborative robot arm with 3D printers or adapted for more complex industrial machine tools. These improvements can support the enterprises to spot potential problems before they occur, optimize performance and reduce costs.