首页> 外文会议>International conference on human-computer interaction >Design Approach for Investigating Multimodal Communication in Dismounted Soldier-Robot Interaction
【24h】

Design Approach for Investigating Multimodal Communication in Dismounted Soldier-Robot Interaction

机译:士兵与机器人交互中多模式通信研究的设计方法

获取原文

摘要

For several decades there has been continuous growth in the field of robotics, with recent trends driving towards a vision of humans collaborating in a cohesive unit with automated counterparts. Enabling true mixed-initiative teaming between a human and robot will require communication capabilities and cognition comparable to human teammates. Multimodal communication is a framework in which interfaces can be created supporting the flexible selection of different modalities (e.g. speech, gestures) for these transactions. A major challenge for human factors researchers investigating human robot collaboration with multimodal interfaces is the current limitations of robots. Therefore, simulations and wizard-of-oz type experiments are heavily employed to measure performance, workload, and other factors in future mixed-initiative scenarios. Although these techniques facilitate experimentation, it can be difficult to transition findings to working prototypes of today's robots. For example, a researcher finds an effective way to convey a robots decision making rationale in a simulation-based study, but has no working robot that can drive the content in reality. Furthermore, the literature regarding multimodal communication with robots applied to the military domain is limited. For example, evaluation of different modalities as part of an interrupting task has been explored in driving scenarios, but not between robot(s) and soldiers. In many cases there is conflict in findings across domains. To address this challenge, this paper describes how a multimodal interface for a real robotic platform developed under the U.S. Army Research Labs Robotics Collaborative Technology Alliance (RCTA) was extended to support standalone simulation of interactions and integration with simulated virtual environments. This functionality enables researchers to assess new interaction techniques using the same software that will interact with a real platform to facilitate transition of their research. Furthermore, experiment design approaches including theory-based tasks in a military relevant mission (Cordon and Search) are discussed.
机译:几十年来,机器人技术领域一直在不断发展,近来的趋势推动着人们对人类与自动对等物紧密协作的愿景的展望。要实现人与机器人之间真正的混合-主动组合,将需要与人类队友相当的沟通能力和认知能力。多模式通信是一个框架,在该框架中可以创建支持这些交易的不同模式(例如,语音,手势)的灵活选择的接口。人为因素研究人员研究具有多模式界面的人类机器人协作的主要挑战是机器人当前的局限性。因此,在未来的混合启动方案中,大量使用模拟和盎司向导类型的实验来衡量性能,工作量和其他因素。尽管这些技术有助于实验,但可能很难将发现转换为当今机器人的工作原型。例如,研究人员在基于模拟的研究中找到了传达机器人决策依据的有效方法,但没有能在现实中驱动内容的工作机器人。此外,有关与应用于军事领域的机器人进行多模式通信的文献有限。例如,已经在驾驶场景中探索了作为中断任务一部分的不同模式的评估,但是没有在机器人和士兵之间进行评估。在许多情况下,跨领域的发现存在冲突。为了解决这一挑战,本文描述了如何扩展由美国陆军研究实验室机器人协作技术联盟(RCTA)开发的真实机器人平台的多模式界面,以支持交互的独立仿真以及与仿真虚拟环境的集成。该功能使研究人员可以使用与实际平台进行交互以促进其研究过渡的相同软件来评估新的交互技术。此外,还讨论了实验设计方法,其中包括与军事有关的任务(Cordon和Search)中基于理论的任务。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号