Fachbereich Informatik - Aktuell
Disputation Johannes Lächele
am Mittwoch, 20. Dezember 2017 um 16 Uhr in Raum A104, Sand 1, EG.
Motion feedback in the teleoperation of Unmanned Aerial Vehicles
Berichterstatter 1: Prof. Dr. Heinrich H. Bülthoff
Berichterstatter 2: Prof. Dr. Andreas Zell
Teleoperation of unmanned vehicles is a valuable tool in scenarios where the operator can not or should not operate the vehicle from on-board. Applications range from hazardous environments where exposure needs to be avoided, control of Unmanned Aerial Vehicles (UAV) to retrieve overviews of inaccessible disaster areas, to deep sea exploration whereon-board operation is simply not possible.
However, limitations in sensor performance, noise and latencies in-troduced in the transmission, and ineffective display of the information to the operator can lead to a reduced amount of information, reduced performance, a loss of situation awareness, and in the worst case a loss of the remote vehicle. The spatial decoupling between the operator and the vehicle is one of the main challenges in teleoperation.
Most setups include one or more control sticks to steer the vehicle and multiple monitors displaying the live video feed of the main vehicle camera and the state information of the remote vehicle. Instead of presenting information purely visually, other feedback modalities can be used to convey vehicle state or information about the task.
The goal of this PhD thesis is to investigate the possibility of providing additional information using motion feedback. Here, motionfeedback is defined as physically moving the operator using a motionsimulator. In the work presented in this thesis a distinction between two motion feedback types is made. Vehicle-state motion feedback de-scribes vehicle motion, while task-related motion feedback is the result of the combination of desired and actual vehicle motion.
The results show that providing motion feedback has a positive effect on performance in teleoperation of remote UAVs. If the remote vehicle is subject to external disturbances, e.g., wind gusts, vehicle-state feed-back showed to improve disturbance rejection capabilities leading to increased performance. Furthermore, motion feedback can be shaped to include additional information about the task with positive effects on performance. This shows that the additional information included in the motion feedback can be used by the operator to improve performance and control behavior.