Project Description: 

As part of a multidisciplinary senior design team, I helped develop the Maestro Arm, a modular wearable robotic arm designed to function as both a wearable third arm and a below-the-shoulder prosthetic. The system features interchangeable mounting configurations and a foot-worn controller that uses machine learning to recognize foot gestures for controlling the arm. The robotic arm was designed  users with temporary or permanent mobility limitations in mind.

My Contributions: 

I led the software development for both the robotic arm and the foot controller. I programmed the microcontrollers, establishing communication between the foot controller and the robotic arm as well as the control logic for the arm's actuators.

For the machine learning component, I designed the neural network architecture, collected and labeled training data, and trained the gesture recognition model, achieving 98% validation accuracy across the six supported foot gestures. I then integrated the trained model into the embedded system, enabling reliable real-time gesture recognition to control the robotic arm.