ROS2 and 3D Vision-Based Automated Pick and Place System for Large-Sized Nuts

This paper presents the design and implementation of a Robot Operating System 2 (ROS 2)- based Pick and Place pipeline designed to detect, pick, and transfer industrial large-sized nuts to a destination box. By leveraging RGB-D camera data, point cloud processing, and Eye-to- Hand calibration, the 3D spatial coordinates and grasp poses of target nuts are extracted in real time. MoveIt2 serves as the primary motion planning framework to ensure collision-free trajectory generation, while action-based ROS 2 communication handles task execution and gripper synchronization. The proposed system was validated through simulation and physical hardware testing, demonstrating high grasp reliability and spatial accuracy.