Arm Sleeve Design Using Electromyogram and MPU6050 Sensor for Robot Arm Control

The use of EMG sensors and IMU sensors in robotics has been widely applied in various fields. A lot of research has been done to create a robotic arm that collaborates with a human arm. However, the challenges that must be faced in reading EMG signals are removing unnecessary noise and smoothing the accelerometer signal readings. Therefore, the sensor system in this study aims to create a smooth EMG signal reading that is able to identify hand movements. The EMG signal will go through hardware filter processes such as low pass filter, high pass filter, notch filter, rectification, and amplification. EMG signal will also be applied to the moving average filter. While the IMU sensor reading will be applied using the DMP feature. From the results of the research conducted, the filter circuit that has been made is able to pass signals in the frequency range of 60 Hz to 1 kHz and is able to identify finger flexion and finger extension movements, as well as when the biceps muscle contracts. Furthermore, the accuracy of the IMU sensor reading for the roll angle value is 98.88%. As for the measurement of displacement from the IMU sensor still needs further research due to the overly sensitive reading of the accelerometer value.