Cheng-Hsuan Li, Kai-Yan Peng, B. Ed., Ling-Tong Wu., and Pei-Jyun Hsieh

Owing to AI advances, more and more software or hardware can be applied for teaching and learning. Moreover, the physical robot was created for delivering learning materials or home care. However, until now, it is tough to apply the real robot in a class. Most robots such as Zenbo or Kebbi Air provide a visualizable and blocky programming tool. In this study, a process integrated with dynamic assessment was proposed to create a teaching robot. The robot includes teaching videos and in-video quizzes with prompts, hints, and direct teaching. In addition, teachers can design some interactive movements from a robot to enhance students' learning interests. In addition, the nonparametric cognitive diagnosis model will be applied to make the robot more personalized for the specific student. Students can be taught according to their non-mastery skills of the teaching content.