A digital twin for 3D path planning of large-span curved-arm gantry robot

• Published in: Robotics and computer-integrated manufacturing 2022-08, Vol.76, p.102330, Article 102330
• Main Authors: Wenna, Wang, Weili, Ding, Changchun, Hua, Heng, Zhang, Haibing, Feng, Yao, Yao
• Format: Article
• Language: English
• Subjects: 3D path planning; Algorithms; Assembly lines; Control methods; Digital twin; Digital twins; Finite element method; Gantry robot; Intelligent manufacturing systems; Multiple objective analysis; NavMesh; Obstacle avoidance; Pallets; Path planning; Robot control; Robots; Strategy; Task scheduling; Trajectory simulation
• ISSN: 0736-5845; 1879-2537
• DOI: 10.1016/j.rcim.2022.102330

•A digital twin platform for the gantry robot is built to realize consistency control.•A grid construction method is proposed for the robot with 3D production environment.•A multi-objective path planning scheme based on assembly line scheduling is proposed. Gantry robots enjoy important applications in unmanned factories and intelligent manufacturing tasks. Aiming at the automated production of customized wooden furniture, this paper proposes a digital twin for three-dimensional (3D) path planning of a large-span curved-arm gantry robot. First, a digital twin platform is built to show the door panel loading and unloading process of the gantry robot. Then the 3D path planning problem and method are proposed based on the built platform. The core of the algorithm consists of a bilateral control method, a 3D mesh construction strategy, and multi-objective 3D path planning using assembly line scheduling. The bilateral control method is used to achieve the perception and control of the gantry robot in the physical space. The mesh construction strategy entails static point set construction, obstacle avoidance analysis, point set construction for leap obstacles, and convex plane construction analysis. Multi-objective 3D path planning is primarily based on the NavMesh algorithm, which leverages an assembly line scheduling model to complete the wooden door processing task. Experimental results show that the proposed method can ensure safety, improved production efficiency, and satisfactory real-time performance in controlling the real gantry robot.