Trajectory Planning and Control of a Robotic Manipulator Using Image Processing and ArUco Markers for Pick and Place Tasks

Abstract

Robotic systems are increasingly being employed in the industry to assist human labor, yet face limitations in movement within the workspace. To explore and develop solutions for unstructured environments, this project aimed to simulate a control system for the robotic manipulator Kuka LBR iiwa 14 R820 with trajectory planning based on image processing and the use of ArUco markers for object detection in the workspace, for pick and place tasks. Therefore, the proportional-integral controller in joint space was selected to adjust joint positions according to the planned trajectory. The computer vision system involves image preprocessing and identification of objects equipped with ArUco markers. The simulation was carried out using the CoppeliaSim software and Python as support, with scenarios featuring objects arranged on a conveyor belt, captured by a camera. Vision system results were satisfactory, with position estimation errors within 0.486 and 6.351 mm, and also rotation between 0.306 and 1.097 degrees. The control responses for joint positions were positive, demonstrating minimal deviations between desired and controlled positions.