Abstract
Industrial equipment, such as dies and moulds, needs inspections in remanufacturing, which is an essential option for implementing circular economy in manufacturing industry. To automate inspection methods such as ultrasonic testing(UT), a robotic arm can be used to inspect with better safety and precision. However, there are some challenges in robot control, for example, orientation control and moving speed adaptive control. In this study, a model-based method will be used to control the orientation of the end-effector of the robotic arm based only on contact forces. A 6 DOF force/torque sensor measures contact force/torque between the end-effector and the surface. A control software platform based on robot operation system (ROS) is established to control the trajectory of robot both in simulation and real-world from scratch. Moreover, a reinforcement learning (RL) algorithm is implemented to optimise the trajectory, that is, the orientation control, the moving speed and position control, during the surface scanning between each waypoint. The proposed method is verified in the simulation environment, and it can be used in real-world robotic UT.
Get full access to this article
View all access options for this article.