Design and Implementation of Upper Limb Controlled by Human Voice and Artificial Intelligence

Given the restrictions and complexity of EMG-based hand prostheses, the present work chose to investigate the possibility of a much simpler method based on voice-operated. Voice commands are simpler to understand than EMG signals to regulate the prosthetic arm. Therefore, the first and most important benefit of this method is that the related prosthesis is simple to use. Speech recognition is a vital focus in artificial intelligence, serving as a prominent mode of human interaction. Researchers have developed speech-driven prosthetic hand systems using traditional speech recognition frameworks and neural network models. This work intends to employ the Raspberry Pi 4 Model B – 4GB RAM embedded inside the prosthetic limb instead of Arduino without requiring a computer to lower hand weight; this offers simplicity of usage. The proposed system captures and transforms speech input into features resembling spectrograms. It processes them using the MLFFNN to categorize speech as signals (words) and forward it to the prosthetic hand, fingers, and wrist control system involving six servo motors. 3D printers created a light and sturdy prosthetic locally. This work stands out for the freedom of usage amputees of