Enhancing Manipulator Flexibility: Real-time Positional Control for Variable Assembly Environments using AprilTag Markers and Edge Detection

Abstract

Most current industrial manipulators are commonly utilized to execute repeated tasks, in which changes in the task require programming of the robot which can be time-consuming. To address this, we provide a method that relays the positional information of the assembly goal position in real-time to a manipulator robot to allow for an assembly process in a variable environment. Our method utilizes AprilTag markers and an edge detection system for high precision location of the assembly goal position in the workspace via sensor integration of 4 RGB cameras. The method also incorporates human-robot interaction mechanism in the form of a digital twin, created in an Isaac Sim environment, to allow for error checking and increase in safety.