EMG-Based Human Machine Interface

Group Members: John Reaser Cochrane, Thomas DiProva, and Jonathan Moron

Advisors: Drs. Yufeng Lu, In Soo Ahn

About Our Project

The project aimed to design and develop a wearable human machine interface device for an in home assistance service robot. The system consists of three main subsystems a EMG signal extraction and classification subsystem, a robot pwm motor controlled service robot, and a live video feedback for user control. The EMG control subsystem utilized a MyoWare muscle sensor board, to amplify the EMG signal, and a Particle Photon microcontroller to implement a artificial neural network (ANN) to classify the EMG signals. The Particle Photon transmitted the commands through WiFi to the Raspberry Pi which then generated PWM specific commands for motor control. The Raspberry Pi 3 also hosted a website which was used for robot navigation and user instructions.


About Us