SBIR Phase II: Personal Service Robotics with Tiered Human-in-the-Loop Assistance

这个小企业创新研究（SBIR）第二阶段项目将创建技术组件、操作界面和集成，以实现实用的个人服务机器人。个人服务挑战机器人在非结构化、杂乱和不断变化的环境中，并且需要执行复杂的任务。我们的新方法允许自主算法在需要时请求人类的帮助。在多个层面上，辅助范围可以从简单地识别部分遮挡的物体，到选择适当的抓取配置，甚至通过直接远程操作来解粘楔形物体；然而，持续的操作人员介入应该很少发生。与传统的远程操作相比，这种方法可以使不完美的自主完成全部任务，同时最大限度地减少体力劳动需求。第一阶段的工作确立了这种联合方法的可行性，同时展示了典型家庭任务的执行。第二阶段的发展将集中于减少人力工作量。主要的创新将来自三个要素：(a)创建和优化不同的操作员界面以满足不同的援助需求，(b)改进自主性以减少援助需求，以及(c)将应用程序和支持算法缩小到相关和期望的服务任务。这些发展将允许单个操作员并行支持多个机器人，具有倍增效应，从而扩大了体力劳动的节省。该项目的广泛影响和商业潜力解决了这样一个事实，即像许多工业化国家一样，美国的人口结构正在向老龄化转变。这意味着劳动力不断萎缩，而支持老年人社区的劳动力需求不断上升，随着老年人年龄的增长，他们在日常生活中需要更多的帮助。机器人通常被认为是扩大有限劳动力供应的一种解决方案。该公司专门针对个人服务机器人，通过家庭护理提供者或辅助生活设施为老年人提供支持。通过减少操作员工作量并允许单个操作员协助多个机器人执行个人服务任务，这种研发工作可以提高服务劳动力的生产率，并提供比传统方式更高质量的个人服务。这种机器人方法也可以更容易地在物理距离上分配劳动力。这将使老年人在熟悉的住房中生活，保持更大的独立性，并减少他们在家中或辅助生活设施的开支。

This Small Business Innovation Research (SBIR) Phase-II project will create the technological components, operator interfaces, and integration necessary to achieve practical Personal Service Robots. Personal service challenges robots with unstructured, cluttered, and ever-changing environments, and requires execution of complex tasks. Our novel approach allows autonomous algorithms to request human assistance as needed. Across multiple tiers, assistance can range from simply identifying a partially occluded object, to selecting an appropriate grasp configuration, or even to unsticking a wedged object by direct teleoperation; however, sustained operator involvement should only occur infrequently. This approach enables imperfect autonomy to complete full tasks while minimizing manual labor needs as compared to classical teleoperation. Efforts during Phase I established the feasibility of this combined approach while demonstrating the execution of typical household tasks. Developments under Phase II will focus on reducing the human workload. Major innovations will occur from three elements: (a) creating and optimizing distinct operator interfaces for different assistance needs, (b) refining autonomy to reduce the need for assistance, and (c) narrowing applications and support algorithms to relevant and desired service tasks. These developments will allow individual operators to support multiple robots in parallel, with the multiplicative effect amplifying manual labor savings.The broader impact/commercial potential of this project addresses the fact that the U.S., like many industrialized nations, is seeing its demographics shift to an older population. This implies a shrinking workforce as well as rising labor needs to support the older adult community, which will require greater assistance with tasks of daily living as they age. Robotics is often considered a solution to stretch a limited labor supply. The company specifically target personal service robotics to support older adults via either home care providers or assisted living facilities. By reducing operator workload and allowing a single operator to assist multiple robots performing personal service tasks, this R&D effort can increase the productivity of service labor and provide a higher quality of personal service than traditionally possible. This robotic approach can also more easily distribute the labor across physical distances. This will allow older adults to age in familiar housing, retain greater independence, and reduce their expenses at home or in assisted living facilities.