Being Doomfist is better than playing Doomfist

Controlling the Overwatch character with electromyography sensors (EMG).

This is a project participating CalHacks - Hello:World

View our progress and gallery at Devpost

Brief Explain

We use the Myoware Muscle Sensor to read the electrical signals from three muscle groups. The analog voltage value is then captured with the Tequila Nano controller board. Then, the 14-bit analog values are send to the USART port alone with the state of a push button.

On the PC side, we use Python to read the data received from the USART port, and then apply a threshold to determine whether a certain action is activated. If the signal is above threshold, we will use win32api and windll to simulate keyboard and mouse events. These events will be registered by Overwatch, thus achieve the successful control of the Doomfist character.

Credits (names in alphabetical order)

Haocheng Yang: Hardware implementation

Yichen Wang: Software implementation

Yifei Li: Software implementation

-T.K.-: Hardware & software implementation