This paper presents a simulation-based approach to robotic object recognition and rearrangement using the KUKA KR5 HW-2 manipulator. The research focuses on developing and implementing advanced algorithms for object detection, classification, and manipulation within a virtual environment. The KUKA KR5 HW-2, a versatile industrial robot, is modelled in the simulation to perform complex pick and place operations. It emphasizes the integration of computer vision techniques with the robot's kinematic model to enhance its ability to recognize various object geometries and efficiently plan rearrangement tasks. The simulation environment allows for rapid prototyping and testing of different scenarios, enabling the optimization of trajectory planning and grip strategies. Our results demonstrate significant improvements in the manipulator's adaptability to diverse object types and arrangement patterns. This work contributes to the advancement of flexible automation in manufacturing and logistics, offering insights into the potential of simulation-based development for robotic manipulation tasks.