INTELLIGENT ROBOTIC SYSTEM FOR AUTOMATIC IDENTIFICATION OF MARKS ON PARTS

The increasing market demands are driving companies to develop intelligent robotic cells capable of making autonomous decisions based on their environment. This work addresses the industrial challenge of distinguishing and tracking marked parts throughout the production process, where precision in identifying these parts and their marks poses a significant obstacle. As a solution, the integration of advanced learning techniques into a vision system for pick and place applications with a parallel delta robot is proposed. The developed system uses cubes with serigraphed letters on their top surface, identified through artificial vision to form words and control the process via a graphical interface. To automate letter recognition, an intelligent model based on a Convolutional Neural Network has been implemented, achieving 100% accuracy. Keywords: Delta Parallel Robot, Artificial Vision, Automation, Neural Networks, Machine Learning.