Method to Estimate Human Inattention in Teleoperation of Mobile Robots

Abstract

In teleoperation of mobile robots the operator is remotely located. As a result, generally the human perception of the remote environment is distorted affecting the mission negatively. Visual information can be degraded because of video images bandwith, time lags, frame rates, point of view and motion effects among other reasons. Although many researchers have proposed a variety of methods for measuring perception, just a few can be used in control closed loop systems. This paper aims to provide a novel metric to the human visual inattention upon risk for a remotely navigated mobile robot. We present both qualitative and quantitative guidelines for designing the metric in a teleoperation of a mobile robot. The method allows to incorporate the metric in a control closed loop system, and task consists in guiding the robot from an initial point to a final one as quick as possible, considering the constraint of avoiding collisions. Furthermore, a haptic cue based on the metric is proposed in order to help the human to avoid collisions. A system stability analysis considering time varying delays is proposed. Additionally, we present a human in the loop experiment of a  teleoperation of a 3D mobile robot simulator in order to remark the advantages of using human factors in the controller.