Automating Quality Control in Traditional Japanese Sweets Packaging: An AI-Driven Approach to Oxygen Indicator Color Detection

This study proposes the development of an AI-based system to automate the color judgment of oxygen indicators used in traditional Japanese sweets. The aim is to reduce human errors and enhance manufacturing efficiency. We introduce a novel approach where AI is utilized to address quality control issues in the packaging process of these sweets. The oxygen indicators, which change color upon interaction with oxygen, are crucial for maintaining product integrity as oxygen can lead to mold growth. Traditionally, the presence of oxygen has been checked visually at the production site, which carries a risk of oversight. Our proposed AI system automates the detection of color changes in the oxygen indicator, thus enabling more accurate and efficient identification of oxygen ingress, thereby improving the overall quality assurance process.