Preliminary Study on the Development of a Closed-Loop Control System Using Optical Feedback

In recent decades, there has been a rapid development of computer vision techniques that have found applications in various engineering fields such as automation and robotics. This topic can also be developed and further explored in servosystems, a field dominated by physical sensors and transducers. This study proposes the development of a closed-loop control architecture with optical feedback starting from an already existing test bench. It suggests replacing the potentiometric transducer used for position feedback with a camera-based system. A primary contribution of this work lies in the selection of components for the development of the system and the creation of a customized dataset for training the object detection algorithm using the YOLOv8 neural network.