Fast robot arm control based on brain-computer interface

Brain-computer interface (BCI) provides a new way to conduct external devices of computer by translating brain signal to computer command, which may help the disables' daily life and enhance the normals behavior capacity. In this paper, a novel visual stimulation-based P300 paradigm was designed to control a robot arm. A camera was used to get the visual feedback of the arm's scene, which enabled users to adjust control signal according to the real condition. Neighbouring stimulation mode based on previous moving orientation was proposed to achieve fast location, which aimed to enhance realtime performance of the robot arm control. Using a novel control logic of fast location, a combination of command choices was devised to carry out gripping and releasing action for the robot arm. Three subjects participated in the experiment. The accuracy in offline training was all above 90.0%. All the subjects achieved the success in controlling the robot arm to move and grip by our BCI stimulation paradigm. The online and offline results showed the effectiveness and speedness of our BCI stimulation paradigm. The result alse indicated the average time saved is 15.01% by the fast location. This study proved the effectiveness of the control logic of fast location to control a robot arm.