Development of a human interface for remote-controlled robots using an eye-tracking system

We have been developing and studying robot/human interfaces, which are required for the smooth operation of remote-controlled robots. When humans control robots from a remote location, it is necessary to have a sense of the environment surrounding the robot. Collecting this information through the human senses of the operator is the key factor in realizing smooth remote control of the robot. This paper describes a method by which the operator gains a sense of the environment surrounding the robot as a 3-dimensional space in the same manner as in normal human sight using an eye-tracking system, namely a method of grasping visual information from the robot using 3-dimensional images. This method is based on the technique of detecting and tracking the line of sight of the operator and then changing the line of sight of the robot in conjunction with it. As a result of using this method to realize a human/robot interface, we were able to demonstrate reduced eye fatigue when viewing 3-dimensional images and to simplify the process of robot control for the operator.