Robot manipulator control using absolute encoder and electromyography signal

Many robots have been being developed based on the characteristics of the human body. Robots that mimic the motion of a human arm have many practical applications. The author built a robot manipulator which is controlled by a human left arm and studied its effectiveness. This robot had two joints and one gripper attached to the end-effector. Two joints worked based on the information from two absolute encoders operating as movement sensors on the operators' left elbow. The gripper picked up and released objects based on the electromyography (EMG) signals from the flexor muscle of left hand. The first joint provided for flexion and extension and the second joint was for supination and pronation rotation. Both of the encoders produced three bits of data resulting in eight combinations. The resolution for the first encoder was approximately 7° and the second was 10.5°. The gripper mode depended on the amplitude of EMG signal. The gripper picked up the object if the signal was bigger than the threshold value and it released the object if the signal was smaller than that threshold value. The threshold used was 100 μν. The result showed that this control design successfully operated the robot joints using human arm movement and the contraction of the flexor muscle controlled the gripper well. This system can be further developed by increasing the resolution of the sensors and flexibility of the mechanical structure.