<![CDATA[The rapid advancement of technology has created opportunities to improve physical education through interactive learning media. This study aimed to develop an interactive learning media application using ML Kit pose estimation to support the learning of fundamental non-locomotor movements among elementary school students. The study employed the Research and Development (R&D) method using the ADDIE model. The product was validated by one learning expert, one subject-matter expert, and one media expert, and implemented with one physical education teacher and 19 elementary school students at MIN 24 Aceh Utara. Data were collected through validation questionnaires, observations, interviews, and system performance testing. Quantitative data were analyzed using descriptive percentage analysis, while qualitative data were used to support product evaluation. The results showed that the developed media was highly feasible, with validation scores of 95% from the learning expert, 93% from the subject-matter expert, and 98% from the media expert. The pose estimation system achieved an accuracy of 99.56% in detecting students' non-locomotor movements. The application provided real-time visual feedback to support movement accuracy and student engagement. In conclusion, the developed interactive learning media is highly feasible and technically accurate for supporting the teaching of fundamental non-locomotor movements in elementary schools. The integration of ML Kit pose estimation provides an innovative approach to enhancing interactive and technology-assisted physical education learning.]]>
