Design and Construction of 4-DOF EMG-Based Robot Arm System

Abstract

Electromyography (EMG) provides an alternative way of providing signal responses from the muscle. As such, the recent trend in developing myoelectric devices has spark the interest in this specific field of study. This is because the traditional controllers lack in certain parts which reduce the utilization of limbs to control devices mainly the robotic arm. However, noise such as crosstalk, motion artifact, ambient noise and inherent noise have become a major issue when handling EMG signals. The preparation of electromyography requires more attention in terms of muscle group selection, electrode placement and condition of the surrounding as it will affect the signal output. The aim of this project is to develop a 4 degree-offreedom (DOF) robotic arm that can be controlled using EMG signals. The correlation between the EMG signal and the robotic arm are required to be identified in order to analyze the performance of robotic arm. Review on the actuator, electromyography methods and microcontroller are done to evaluate the techniques used from past researches. The methods of this project include hardware development of robotics arm, development of forward kinematic, sensor calibration and electrode positioning and experiment on classification and validation of EMG signals based on hand gestures. The experiment showed that the sampling rate and arm position affect the EMG signal output. In addition, the controllability of the robotic arm was low because the motors are controlled independently. The objective of the project has been achieved as the EMG-controlled robotic arm has been successfully developed. The robotic arm is still available for improvement by adding multiple channel sensors and implementing a wireless system.

Comparison of two Wireless Electromyography Sensor Module Designs using wet electrodes and dry electrodes at the time of Sitting motion to stand

“…Researchers used the Myoware Muscle Sensor for this project because it was able to measure muscle activation in terms of electrical potential. The sensor can produce two types of output, namely amplified raw electromyography and an electromyography envelope [29].…”

Detection Signal Electromyograpy using Dry Electrode and Disposible Electrodes on the Upper Extremity when Lifting Weights

“…Because the Myoware muscle sensor could assess muscle activity in terms of electrical potential, researchers used it for this purpose. In this case, the sensor had two output options: envelope electromyography and amplified raw electromyography signal [26].…”