AI画像解析結果を説明変数に用いたインフォマティクスによる粒子製造の最適条件の探索

抄録

In the manufacturing industry, there is a growing focus on the application of AI-driven material informatics (MI) and process informatics (PI). However, issues remain in the handling of information that is difficult to convert into data, particularly with regard to image data. While images contain critical information related to material performance, conventional methods have difficulty in accurately analysis. To address these issues, this paper introduces AI technology capable of quantifying images and presents examples that immediately quantify parameters related to particle size, shape, and color. Furthermore, we report on the optimization of manufacturing conditions through the utilization of this quantifiable image information, a capability that was previously unattainable.