Humanoid Robot Operation by a Brain-Computer Interface

Brain-computer interface (BCI) provides a novel communication channel between human and external devices, which translates the brain activities directly to the computer commands. In this paper, an event-related potential (P300) based BCI paradigm was constructed to control a real humanoid robot. First, the images of the robot and the environment were integrated with the P300 graphic stimulation interface. Then the subjects could obtain all the necessary feedback information in one computer screen. Second, a variable-pace mechanism was designed for human-robot cooperating in different situations. Third, an on/off button based on two-layer P300 decision was devised, allowing the subject to turn on/off the system. Three subjects participated in this study, and the accuracy in offline test was above 90.0%. Furthermore, all of the subjects successfully controlled the humanoid robot to accomplish a walk task with our BCI paradigm. The results prove the effectiveness of our improved BCI paradigm to control a humanoid robot. In the meanwhile, this study also demonstrates that it can be used as a potential technology applied to rehabilitation, aerospace and other related applications.