查看文章

This paper presents a system that can be implemented in brewing factories using actual facilities as a testbed. The system performs Machine Learning based vision classification with images collected by machine vision. It requires a large amount of computational power, but edge devices in factory generally do not support powerful processing capabilities. Computing resource limitations inherent in edge devices can be overcome by using a centralized server. We show the demonstration in a actual factory and accommodate the necessary delays in edge devices while obtaining the necessary data.