Shadow Mode Robotic Arm

This paper presents the design and implementation of a Shadow Mode Robotic Arm, a high- precision teleoperation system designed to replicate human arm movements in real-time. The system leverages an ESP32-CAM module for vision-based pose estimation, utilizing the MediaPipe framework to track key upper-body landmarks. The extracted joint coordinates are processed and mapped to a multi-degree-of-freedom (DOF) robotic arm actuated by MG995 metal-gear servos. An ESP32microcontroller serves as the secondary control unit, generating Pulse Width Modulation (PWM) signals for synchronized actuation. The system features a dual-mode operation: a direct mimicry \\\"Shadow Mode\\\" and a remote web- dashboard control mode. Experimental results demonstrate a low-latency response (<100ms) and high kinematic fidelity, offering a cost-effective solution for hazardous environment handling, telemedicine, and industrial automation.