Abstract
This research aims to develop a BMI for controlling a mobile robot in road-like environments. The robot makes a local map of the surrounding scene using a laser range finder and recognizes the intersection type, which is sent to the BMI. The BMI flashes LEDs corresponding to possible branches of the current intersection, and determines the operator's choice using steady-state visual evoked potentials (SSVEP). The choice is sent back to the robot for generating an automous behavior such as a right turn. Commnunication between the robot and the BMI is performed through a wireless LAN. The robot also sends a video stream of a camera put on a robot for the operator being able to recognize the scene. We carried out online experiments and validated that we can control the mobile robot using the BMI.
