Sensitivity of Eye-Tracking Measures to Variations in Mental Workload while Learning to Operate a Physically Coupled Robot
Sensitivity of Eye-Tracking Measures to Variations in Mental Workload while Learning to Operate a Physically Coupled Robot
复制标题
学习操作物理耦合机器人时眼动追踪测量对脑力负荷变化的敏感性
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
D. Srinivasan
中科院分区：
文献类型：
--
作者：
Satyajit Upasani;Qi Zhu;E. Du;A. Leonessa;D. Srinivasan
协作机器人（cobots），特别是那些物理耦合的机器人，例如假肢，外骨骼和辅助机器人，与旧的工业机器人相比，已经变得更可用，更通用，更安全，可以与人类近距离操作（hadadin & Croft, 2016）。然而，最近的研究表明，协作机器人仍然会给用户的注意力和认知运动资源带来巨大的工作量（Marchand等人，2021；Stirling等人，2020），并且可能需要时间和精力来学习（Aronson等人，2018;Cornwall, 2015）。在这种情况下，重要的是要了解控制这些复杂设备所涉及的认知挑战，以及不同操作员有效利用它们的学习/培训需求。眼动追踪是一种很有前途的测量心理工作量的技术，因为它可以提供各种瞳孔和凝视行为指标（Holmqvist等人，2011），并且对认知和运动技能学习敏感（Foroughi等人，2017;Sailer, 2005）。我们的研究旨在量化在不同任务难度下学习使用协作机器人的过程中眼球追踪指标的变化。参与者在虚拟现实（VR）的基础上完成了一项手动物体拾取和放置任务，同时在物理上操纵百特协作机器人
Collaborative robots (cobots), especially ones that are physically-coupled, e.g. prostheses, exoskeletons and assistive robots, have grown much more usable, versatile, and safer to operate in close proximity with humans compared to older industrial robots (Haddadin & Croft, 2016). However, recent research suggests that cobots can still impose a significant workload on the user’s a ttentional and cognitive-motor resources (Marchand et al., 2021; Stirling et al., 2020) and may require time and effort to learn (Aronson et al., 2018; Cornwall, 2015). In this context, it is important to understand the cognitive challenges involved in controlling these complex devices, and the learning/training needs for diverse operators to effectively utilize them. Eye-tracking is a promising technique for measuring mental workload, since it can provide a variety of pupillary and gaze-behavioral metrics (Holmqvist et al., 2011), and is sensitive to cognitive-and motor-skill learning (Foroughi et al., 2017; Sailer, 2005). Our study aimed to quantify the changes in eye-tracking metrics over the course of learning to use a cobot under varying levels of task difficulty. Participants performed a bimanual object pick-and-place task based in virtual reality (VR), while physically manipulating a Baxter collaborative robot