Development of an Intelligent Automated Product Accounting Tool in Production

Abstract

In this article, the authors consider the process of developing an intelligent automated tool for real-time accounting of products in production, characterized by a large flow of products on the conveyor. Special attention is paid to its use in bakery production, where high accuracy and speed of product counting are critically important for ensuring the efficiency and quality of production processes. The main goal of the research is to create an automated system that will use advanced computer vision techniques and machine learning algorithms to count products on the assembly line in real time. It is proposed to use a modern neural network algorithm based on the YOLOv8 architecture. This will significantly reduce the human factor, improve data accuracy and optimize production processes, and automate product reporting. The architecture of the system includes an IP camera and a microcomputer. The rationale for the choice of equipment is given, calculations for all the necessary camera settings and parameters and equipment requirements, as well as performance tests, and a report on the quantity of manufactured products is presented. The key stages and models used at the application design stage are described: a database model, a sequence diagram, and a use case diagram.The rationale for the choice of system development technologies is also given. In the future, it is proposed to reconfigure this system to recognize the quality of products, in particular, burnt or unformatted bread. The described solution is another proposal of our engineering team as part of the ongoing work on the integrated automation of the bakery in accordance with modern requirements for the transition of production processes of enterprises to the principles of Industry 4.0.