Implementation of a Pick-and-Place Robotic System Using Franka Arm and Vision-Based Object Detection: Technical report - 2025

Abstract

This report presents the design, implementation, and evaluation of a pick-and-place robotic system using the Franka Emika robotic arm and a Stereolabs ZED2 camera for object detection and localization. The system employs 3D depth sensing through the ZED2 camera, mounted on the robot’s end-effector, to provide real-time 3D position data of the target object. These coordinates are transformed into the robot’s base frame using a calibrated transformation matrix, enabling precise manipulation. The motion planning is executed using the Robot Operating System (ROS) framework, ensuring smooth and collision-free trajectories. The system dynamically updates the object’s position during the pick-and-place sequence, allowing for accurate handling even in cases of displacement. Experimental results demonstrate the system’s capability to execute high-precision pickand-place tasks, highlighting its suitability for advanced applications such as precisionassembly, warehouse automation, and medical robotics.