Robotic system with Twincat Vision: techniques for identifying and manipulating geometric objects on a conveyor belt

The integration of computer vision in Industry 4.0 marks an important revolution. Many industrial companies have adopted machine vision systems to identify defective products, assembly verification, robot guidance and OCR reading; arduous tasks that were previously labor-intensive, error-prone and time-consuming when performed by human. These advanced machine vision systems, equipped with cameras and artificial intelligence, now capture and analyze image of objects with remarkable precision, handling different functions and significantly enhancing operational accuracy. In fact, vision systems play a crucial role in robotics, enabling machines to operate in unstructured environments. This thesis investigates the application of TwinCAT Vision library for the detection of geometric objects. TwinCAT Vision library facilitates industrial image processing tasks such as the detection, identification or measurement of objects directly within the PLC in real-time. It supports the use of images already saved on the PC as well as the deployment of a GigE Camera for real-time operations. Utilizing the camera MakoG192B, TwinCAT Vision identifies various shapes in motion and accurately locates their center coordinates. This capabilities enhances robot interactions with production lines, which are often controlled by PLCs. This study highlights different vision algorithms for shape center detection and demonstrate their effectiveness through rigorous experimental tests. These algorithms enable precise localization of objects centers, which is crucial for tasks requiring high accuracy. The coordinates determined by the system are utilized by a SCARA robot for manipulation, thereby proving the strategy’s efficacy in real-world applications. By integrating image processing into the TwinCAT platform, the system achieve highly synchronized control applications and extremely short response times, showcasing significant advancements in both the efficiency and reliability of automated industrial processes.