Push-Up Movement Detection and Counting System Using MPU6050 Accelerometer Sensor Based on Arduino

This study presents a real-time push-up movement detection and quality evaluation system using the MPU6050 accelerometer sensor on the Arduino platform. The system integrates advanced algorithms, including Kalman filtering, to enhance sensor accuracy and reduce noise in body position tracking. Comparative analysis shows that the Kalman filter effectively smooths raw sensor data, ensuring precise detection of upward and downward phases during push-ups. From device testing, discrepancies between manual and device observations were recorded, with an average discrepancy of 3 points for the first respondent, and 1–2 points for the others. These discrepancies were due to rapid movements when respondents reached their maximum push-up points, affecting the device's responsiveness. Despite these differences, the system demonstrated an overall accuracy of 81.5%, validating its effectiveness. The approach improves push-up detection and repetition counting, making it suitable for real-time fitness applications for cadet at the Indonesian Air Force Academy.