Controller-Driven Hand-Eye Vision System for KUKA LBR iiwa Robot

Computer vision is a crucial component of many industrial robotic applications, as it allows one to cope with dynamic environments. It is common, especially in the case of computer vision tasks, to delegate the vision task to an external control unit which processes the image and exchanges the results of the image processing procedure with the robot controller. The proposed work presents an alternative approach by integrating a computer vision system into the KUKA iiwa ecosystem, effectively eliminating the need for any additional hardware component other than the camera itself. The applicability of the proposed approach is confirmed by the adoption of commercially available hardware and open-source software tools. A standard blob analysis algorithm has been implemented in the controller together with the acquisition algorithm, and experimental tests have proved the effectiveness of the system. An HTML-based user interface has been developed as well to provide visual feedback to the user, for easier troubleshooting and easier system setup.