We have tried to show that EMG signals can be classified in real-time with an extremely high degree of accuracy for controlling a robotic arm-and-gripper. We studied an offline analysis of an action classification problem based on EMG signals for different subjects as a function of the number of recording sites (electrodes) used for classification. We then demonstrated that the proposed method allows to use EMG signals to efficiently solve several reasonably complex real-time servo motor tasks like a robotic arm, and pick and- drop movements using a 5 degrees-of-freedom robotic arm.
Our project demonstrates that limb muscles contain enough information to reliably distinguish between a large numbers of actions; the interpretation of these actions is left to the user. Thus, for control of other prosthetic devices, one has the option of customizing the actions to better suit the device in question and the desired control. This customization has to be done on a case-by-case basis