Investigation of Inverse Kinematics Algorithms For Trajectory Resolution In Ros-Based Cobot Pick-and-Place Systems

Industry 4.0 is driving automation to increase efficiency and precision in production processes. Collaborative robots (Cobots) have stood out for their ability to perform repetitive tasks quickly and flexibly, being easily pro-grammable for different demands. When integrated with technologies such as IoT and artificial intelligence, they become even more versatile, reducing costs, increasing quality, and minimizing the need for human intervention in dangerous or routine tasks. This paper presents the development of a system aimed at solving inverse kinematics in pick-and-place robots, emphasizing numerical methods such as gradient descent and BFGS ((Broyden-Fletcher-Goldfarb-Shanno). The system com-bines a Cobot, the robot operating system (ROS), and a perception module for organizing parts in a simulated factory environment. The architecture de-veloped allows different kinematics approaches to be explored, highlighting the balance between precision, efficiency, and ease of implementation. The results show that, although analytical methods have higher computational performance, numerical methods offer greater flexibility and simplicity of application and are entirely suitable for solving inverse kinematics. This study reinforces the viability of numerical methods in solving inverse kine-matics problems to be adopted in some tasks in industrial automation.