Human–Robot Interaction in the Context of Simulated Route Reconnaissance Missions

Abstract

The goal of this research was to examine the ways in which human operators interact with simulated semiautonomous unmanned ground vehicles (UGVs), semiautonomous unmanned aerial vehicles (UAVs), and teleoperated UGVs (Teleop). Robotic operators performed parallel route reconnaissance missions with each platform alone and with all three platforms. When given all three platforms, participants failed to detect more targets than when given only the UAV or UGV; they were also less likely to complete their mission in the allotted time. Target detection during missions was the poorest with the Teleop alone, likely because of the demands of remote driving. Spatial ability was found to be a good predictor of target-detection performance. However, slowing sensor feed video frame rate or the imposition of a short response latency (250 ms) between Teleop control and Teleop reaction failed to affect target-detection performance significantly. Nevertheless, these video image manipulations did influence assessment of system usability.