Design and implementation of a new teleoperation control mode for differential drive UGVs

Abstract

PostprintThis is the accepted version of a paper published in Autonomous Robots. This paper has been peerreviewed but does not include the final publisher proof-corrections or journal pagination.Citation for the original published paper (version of record):Ögren, P., Svenmarck, P., Lif, P., Norberg, M., Söderbäck, N. (2014) Design and implementation of a new teleoperation control mode for differential drive UGVs. Abstract In this paper, we propose and implement a new control mode for teleoperated Unmanned Ground Vehicles (UGVs), that exploits the similarities between computer games and teleoperation robotics.Today, all teleoperated differential drive UGVs use a control mode called Tank Control, in which the UGV chassis and the pan tilt camera are controlled separately. This control mode was also the dominating choice when the computer game genre First Person Shooter (FPS) first appeared. However, the hugely successful FPS genre, including titles such as Doom, Half Life and Call of Duty, now uses a much more intuitive control mode, Free Look Control (FLC), in which rotation and translation of the character are decoupled, and controlled separately.The main contribution of this paper is that we replace Tank Control with FLC in a real UGV. Using feedback linearization, the orientation of the UGV chassis is abstracted away, and the orientation and translation of the camera are decoupled, enabling the operator to use FLC when controlling the UGV. This decoupling is then experimentally verified.The developments in the gaming community indicates that FLC is more intuitive than Tank Control and reduces the well known situational awareness problem. It furthermore reduces the need for operator training, since literary millions of future operators have already spent hundreds of hours using the interface.

A New UGV Teleoperation Interface for Improved Awareness of Network Connectivity and Physical Surroundings

“…The new user interface, as most robot teleoperation UIs, is composed of two parts: receiving control commands from the operator; and providing feedback to the operator. In the former, we use a gamepad controller and the FLC interface [26] and in the latter we present the DoA as an extra sensory feedback in addition to the video stream. The hardware configuration and associated signal processing required to realize the new interface is presented below.…”

“…Carefully integrating the DoA information into the visual feedback is crucial. For this we use FLC (Free Look Control) [26] as the control layer. FLC is essentially a "navigateby-camera" mode as envisioned in [27].…”

Development and analysis of an operator steering model for teleoperated mobile robots under constant and variable latencies

“…A gamepad was chosen for use in the study rather than a steering wheel or other input devices such as multi-degree-of-freedom haptic devices because of its widespread usage in mobile robot teleoperation systems. 40,41 A small amount of noise was artificially added to the gamepad input, generated from a uniform distribution on the range of [−10%, 10%] of the maximum possible input command, to simulate the noise present in a physical robot system. Because the focus of this study is to investigate steering behavior rather than combined longitudinal and lateral control, operators do not have control over the robot speed.…”

Adaptive object centered teleoperation control of a mobile manipulator

“…One way of using inspiration from computer games was presented in [10], where the classical robot control mode of Tank Control was replaced with Free Look Control which is used in many computer games.…”

“…• Keep desired distance from object, (7) • Keep desired orientation w.r.t object, (8) and (9) • Limit minimum end-effector altitude, (10) • Avoid collision between robot and object, (11) • Move according to user input. (12) Formally, the constraints can be stated as …”