The rapid evolution of digital technologies has opened new opportunities for transforming vocational education, particularly in CNC (Computer Numerical Control) machining. Despite its vital role in preparing skilled workers, CNC instruction in Indonesian vocational schools remains constrained by teacher-cantered methods, static materials, and limited access to costly machines, resulting in low engagement and achievement. This study aimed to design, implement, and evaluate Android-based Augmented Reality (AR) learning media for CNC TU-2A machines to enhance students’ cognitive performance, psychomotor skills, and classroom participation. Using a Classroom Action Research (CAR) model across two cycles, the research involved 31 eleventh-grade Mechanical Engineering students at SMK Negeri 5 Padang. Data were collected through cognitive tests, student activity observations, and surveys. Results showed significant improvements: average cognitive scores rose from 75.91 to 82.47, classical mastery increased from 54.83% to 100%, and psychomotor scores improved by 3.24 points. Student learning activities also climbed from 72.2% to 80.4%, with discussion and collaboration showing the highest gain (17%). While barriers such as device limitations and technical issues arose, they were addressed through device sharing, offline content, and teacher mentoring. Findings confirm AR as an effective, scalable tool for modernizing CNC instruction, fostering engagement, and preparing vocational students for Industry 4.0 learning demands.
                        
                        
                        
                        
                            
                                Copyrights © 2025