Phoneme-based speech recognition for commanding the robotic Arm

Controlling a robot still requires traditional user interfaces. A more intuitive approach is using verbal or gesture commands. In this paper, we propose a robotic arm that can recognize the human voice commands. The Speech recognition is an essential asset for the robot, enhancing its ability to interact with human beings using their natural form of communication. The approach is verified by deploying the robotic arm into different environments with low, high and medium noise perturbations, where it is tested to perform a set of tasks. With this approach, we have successfully reduced the response time, enhanced the accuracy of the robotic arm to grasp the voice commands and reduced the sentence overlapping by a significant amount. The entire system is divided into three modules: the manipulator, the voice recognition module and the microcontroller. The robotic arm is programmed to orient to the direction where the signal to noise ratio is maximum.