Applied AI with PLC and IRB1200

Abstract

This paper presents the application of artificial intelligence applied to image classification according to CNN (Convolutional neutral network) algorithm. Two AI (artificial intelligence) models were submitted for verification. The first was based on mobilenet algorithm, and the second was based on the application of several layers of Convolution, ReLU and Maxpooling. After examining the AI, a solution using image classification was proposed for the IRB1200 robot, which encircles symbols assigned to color saturation on the detected colors. In the introduction, articles were selected that are related to the application of AI in various industries. The second section describes mathematical relationships for algorithms used in the project for image classification. Next, a multi-task control system for a robot is described, which performs specific tasks based on image verification by CNN-based artificial intelligence. The AI model and system handling reading of the image from a camera were submitted for verification, based on the tasks to be performed by the IRB1200 robotic arm based on color recognition. Verification was carried out in the head office of a company called Sii.