NRI: Socially Aware, Expressive, and Personalized Mobile Remote Presence: Co-Robots as Gateways to Access to K-12 In-School Education

参与学校环境对儿童的社会、情感和认知发展和学习至关重要。长期以来，人们已经认识到，学生在校经历的质量不仅对学业和成就结果很重要，而且对培养自尊、自信和总体心理健康也很重要。然而，每年有26.6%的美国儿童面临健康或行为挑战，导致他们大量缺课，13%的美国K-12公立学校学生因学习障碍或情绪障碍而接受干预。该项目的重点是利用移动远程协作机器人作为一种手段，为许多不能上学的K-12岁儿童提供对他们的发展和未来结果至关重要的课程和社会学习体验。对于在家的学生来说，使用移动远程呈现进入K-12教室可能是将物理隔离与学校环境的影响降至最低的有效途径。该项目开发的方法能够创建个性化的机器人，允许在动态的K-12班级环境中共享自主性、适合社会的运动和具有社会表现力的非语言交流，使儿童能够真正体现在课堂上，即使是从远处。这个NRI项目的影响覆盖了整个K-12教育，但也适用于课堂环境之外的移动远程呈现系统的一般用途，用于教育和培训。此外，该项目将研究主题与外联联系起来；它让K-12年级的学生和教师参与合作机器人主题的活动，并在大型公共场所举办年度NRI主题研讨会。更广泛的推广计划旨在培训STEM学生，这样他们不仅可以成为机器人和其他技术的最终用户，而且能够自己开发这些技术，从而为美国STEM劳动力做出贡献。这项提议的重点是开发移动远程呈现(MRP)联合机器人系统的控制算法，该算法将改善人类进入学习/培训环境的机会，重点是在家的K-12学生，但对所有年龄段和各种背景的用户具有普遍影响。与MRP系统的合作已经确定了促进自然远程交互和学习所必需的关键技术能力：1)简单的、适合社会的自主行为和上下文感知，以减少用户的认知负荷；2)传达用户的情感和交流意图的表现力；以及3)通过MRP个性化用户交互方式的能力。该项目通过参与式用户知情算法开发、系统集成和评估来应对这些挑战。具体地说，它首先开发了一种方法来自动化和促进操作员和MRP的空间和社会背景感知，并使用它来支持两个研究推动力，社会适当性和表现力，以及用于个性化两者的算法方法。为了在选定的真实世界环境中得出结果，在K-12课堂环境中执行迭代设计和评估，涉及不同年龄和教育程度的用户，提供对协作机器人的相关性、有效性和健壮性的测试。该项目汇集了两位具有过去成功合作记录的跨学科专家和三个合作伙伴：行业、部署和外联，致力于制定具有具体可评估里程碑的项目时间表。

Participating in the school environment is essential to children's social, emotional, and cognitive development and learning. It has long been recognized that the quality of a student's school experience is important not only for the academic and achievement outcomes, but for fostering self-esteem, self-confidence, and general psychological well-being. Yet annually 26.6% of America's children have health or behavioral challenges that cause them to miss significant amounts of school, and 13% of all US K-12 public school students receive interventions due to learning disabilities or emotional disturbances. This project focuses on the problem of using mobile remote presence co-robots as a means to provide numerous K-12 aged children who cannot be present in school access to the curricular and social learning experiences critical to their development and future outcomes. Using mobile remote presence for access to K-12 classrooms for homebound students may be a powerful gateway for minimizing the effects of physical separation from the school environment. This project develops methods that enable the creation of personalizable robots that allow shared autonomy, socially appropriate movement and socially expressive nonverbal communication in dynamic in-class K-12 environments, allowing children to be truly embodied in the classroom, even from a distance. The impact of this NRI project spans K-12 education at large, but also applies to general uses of mobile remote presence systems outside of the classroom setting, for both education and training. In addition, the project connects the research themes with outreach; it engages K-12 students and teachers in co-robot-themed activities and holds annual NRI-themed workshops at large-scale public venues. The broader outreach program is designed to train students in STEM, so they can become not only end users of robotics and other technologies but capable of developing such technologies themselves, thereby contributing to the US STEM workforce. This proposal focuses on developing control algorithms for mobile remote presence (MRP) co-robot systems that will improve human access to a learning/training environment, focusing on homebound K-12 students, but with general implications to users of all ages and a variety of contexts. Work with MRP systems has identified key missing technical capabilities necessary for facilitating natural remote interaction and learning: 1) simple, socially-appropriate autonomous behavior and context awareness that reduces user cognitive load; 2) expressiveness for conveying the user's affect and communicative intent; and 3) the ability to personalize the way the user interacts through the MRP. This project addresses these challenges with participatory user-informed algorithm development, system integration, and evaluation. Specifically, it first develops an approach to automating and facilitating spatial and social context awareness for the operator and the MRP, and uses it to enable the two research thrusts, social appropriateness and expressiveness, with algorithmic methods for personalizing both. To ground the results in the selected real-world context, iterative design and evaluation is performed in the K-12 in-class setting, involving users across the age and education span, providing a test of the co-robot's relevance, effectiveness, and robustness. The project brings together a pair of interdisciplinary experts with a track record of successful past collaborations and three partners: industry, deployment, and outreach, committed to a project timeline with specific evaluable milestones.