Smart Quality Management System at the Manufacturing Sector using Deep Learning for Anomaly Detection

Abstract

An assembly line is a place where materials are put together and processed to check if the end product meets the quality standards. One of the biggest challeng-es faced by the manufacturing sector is quality management. Any newly built product is prone to have surface defects like scratches or dents or paint errors which might happened during the process of manufacturing and transportation. Currently such quality checks are being done manually using Human Vision. Now with the advent of IoT and Deep Learning techniques, we could build an anomaly detection system that recognize defects on the surface of the system by capturing photographs of the Automobile in the assembly line and sends it to an image processor system for validation. We were able to detect defects with 90% and above accuracy with the existing detection algorithms.