This project presents a robotic arm simulation controlled via natural language, utilizing a Large Language Model (LLM) to convert spoken commands into structured motion instructions. Spoken phrases such as “move up by 10 centimeters” are parsed into executable JSON commands, enabling real-time 3D movement of a virtual robotic arm. The simulation integrates speech recognition, natural language processing, and forward kinematics, providing an intuitive interface for exploring multimodal AI-based control systems. 

An interactive simulation of a 3D robotic arm controlled by real-time hand tracking via webcam. This system uses MediaPipe and OpenCV to detect hand positions and interpret directional gestures (left, right, up, down) as movement commands. A graphical interface built with Tkinter and Matplotlib visualizes the arm’s response, enabling intuitive, vision-based control for human-robot interaction.