Application of Automated Quality Control in Smart Factories - A Deep Learning-based Approach

Abstract

Industry 4.0 is the ongoing automation of conventional manufacturing and industrial applications using smart technology. Quality control (QC) is a set of procedures to ensure that a manufactured product adheres to a defined set of quality criteria or meets the requirements of the customer. Many applications within the manufacturing domain employ image-processing or machine learning systems but deep learning-based applications are rare. The goal of this project is to leverage deep learning methods for the automation of quality control. A visual QC automation application is proposed that utilizes a camera placed over a product assembly line containing 3-D printed product samples in a smart factory prototype setup for data collection. After model training, the model will perform object detection and recognition for analyzing complex free-form products and perform product dimension and surface analysis to identify the products that meet the quality control guidelines.