当前位置: X-MOL 学术PRESENCE Virtual Augment. Real. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Effects of Throughput Delay on Perception of Robot Teleoperation and Head Control Precision in Remote Monitoring Tasks
PRESENCE: Virtual and Augmented Reality ( IF 0.7 ) Pub Date : 2020-02-01 , DOI: 10.1162/pres_a_00328
Jason Orlosky 1 , Konstantinos Theofilis 2 , Kiyoshi Kiyokawa 3 , Yukie Nagai 4
Affiliation  

For robot teleoperators, replicating head motion in a remote environment is often necessary to complete monitoring and face-to-face service tasks. Especially in the case of a stereoscopic camera rig, remote cameras need to be moved to match the operator's head position and optical axes in order to interact with a remote entity, follow targets, or reorient. However, mechanical, computational, and network delay in such teleoperation can cause intersensory conflict, perceptual deficits, and reduction in performance, especially during activities that require head rotation. In this article, we evaluate the effects of view reconstruction on performance and perception for remote monitoring tasks. To do so, we first implemented a panoramic reconstruction method to reduce perceived latency in a humanoid robot, which allows us to compare and contrast latency. Next, we designed a bidirectional remote control system for the robot and set up a series of experiments where participants had to conduct focused head control tasks through the perspective of the robot. This allowed us to compare the effects of latency on head movement, accuracy, and subjective perception of the interface and remote teleoperation. Results showed that panoramic reconstruction significantly improved perception and comfort during teleoperation, but that performance only improved for tasks requiring slower head movements.

中文翻译:

远程监控任务中吞吐量延迟对机器人遥操作感知和头部控制精度的影响

对于机器人远程操作员来说,在远程环境中复制头部运动通常是完成监控和面对面服务任务所必需的。尤其是在立体摄像机装备的情况下,需要移动远程摄像机以匹配操作员的头部位置和光轴,以便与远程实体交互、跟踪目标或重新定向。然而,这种远程操作中的机械、计算和网络延迟会导致感觉间冲突、知觉缺陷和性能下降,尤其是在需要头部旋转的活动中。在本文中,我们评估视图重建对远程监控任务的性能和感知的影响。为此,我们首先实施了一种全景重建方法,以减少类人机器人的感知延迟,这使我们能够比较和对比延迟。接下来,我们为机器人设计了双向遥控系统,并设置了一系列实验,参与者必须通过机器人的视角进行集中的头部控制任务。这使我们能够比较延迟对头部运动、准确性和界面和远程远程操作的主观感知的影响。结果表明,全景重建显着改善了远程操作过程中的感知和舒适度,但仅在需要较慢头部运动的任务中才能提高性能。这使我们能够比较延迟对头部运动、准确性和界面和远程远程操作的主观感知的影响。结果表明,全景重建显着改善了远程操作过程中的感知和舒适度,但仅在需要较慢头部运动的任务中才能提高性能。这使我们能够比较延迟对头部运动、准确性和界面和远程远程操作的主观感知的影响。结果表明,全景重建显着改善了远程操作过程中的感知和舒适度,但仅在需要较慢头部运动的任务中才能提高性能。
更新日期:2020-02-01
down
wechat
bug