I-Corps: Healthcare Service Robots

过去机器人的商业化努力一直由移动基地主导。机械臂往往待在笼子里或远离人类（例如，制造业）。这个I-Corps团队希望将移动机械手带到人类环境中，以执行人们不喜欢的重复性手动任务。研究小组最初的假设是，表面清洁就是这样一项任务。预计表面卫生的自动化可以实现更高频率的表面覆盖，创造一个更卫生的环境，减少疾病。这样的机器人也有可能自动化不受欢迎的体力劳动，以及减少与重复性体力劳动相关的人类伤害。将需要较低技能或体力劳动的任务自动化，可以让人类更多地专注于高技能劳动，从而改善劳动力短缺问题。因此，这个项目旨在定义智能移动机器人的商业应用。无处不在的、友好的、可以被更多人使用的机器人的出现，也将理想地激励年轻一代追求STEM教育，从而形成一支技能更强、更具竞争力的美国劳动力队伍。本项目旨在定义具有手臂的智能移动机器人（移动机械手）的商业应用。基于8年多的人机交互研究成果，该团队打算应用演示学习（LfD）算法，使没有机器人经验的最终用户可以对机器人进行编程，这可以通过直观的社交对话来完成。机器人还需要在人类面前以透明和可预测的方式自主行动，这需要开发具有社会意识的操纵和导航行为。在I-Corps项目期间，团队的大部分方法将致力于客户发现的访谈，最初的目标客户是医疗机构。为了验证这一细分市场，该团队打算访问医院和长期护理机构，以确定客户的痛点，以及采用机器人技术的流程和障碍。此外，该团队计划采访专家，参加会议，并进行文献检索，以更好地了解该领域；并且，在I-Corps项目结束时，它预计已经定义了一组机器人必须执行的关键任务，以为我们的目标客户群提供价值。这些任务将定义机器人硬件和软件的最小可行设计规范。

Past commercialization efforts for robotics have been dominated by mobile bases. Robotic arms tend to stay in cages or far from humans (e.g., manufacturing). This I-Corps team wants to bring mobile manipulators to human environments to perform repetitive manual tasks that people find undesirable. The team's initial hypothesis is that surface cleaning is one such task. It is anticipated that automation of surface sanitation can achieve more surface coverage with higher frequency, creating a more sanitary environment that reduces illness. Such a robot also has the potential to automate undesirable manual labor, as well as reduce human injuries associated with repetitive physical work. Automating tasks requiring lower skills or manual labor could ameliorate the labor shortage by allowing humans to focus more on high-skilled labor. This project, therefore, aims to define a commercial application for an intelligent mobile robot. The presence of ubiquitous, friendly robots that can be used by more people will ideally also inspire younger generations to pursue STEM education, leading to a more skilled and competitive American workforce. This project aims to define a commercial application for an intelligent mobile robot with an arm (mobile manipulator). Based on over eight years of research results in Human-Robot Interaction, the team intends to apply learning from demonstration (LfD) algorithms to make the robot programmable by end users with no robotics experience, which can be accomplished using an intuitive social dialogue. The robot will also need to act autonomously in the presence of humans in a way that is transparent and predictable, which requires developing manipulation and navigation behavior with social awareness. The bulk of the team's methodology during the I-Corps program will be dedicated to interviews for customer discovery, with an initial target customer as health care establishments. To validate this segment, the team intends to visit hospitals and long-term care facilities to identify customer pain points, as well as processes and barriers for adopting robotic technology. In addition, the team plans to interview experts, attend meetings, and conduct literature searches to better understand the domain; and, by the end of the I-Corps program, it expects to have defined a set of key tasks that the robot must perform to provide value to our target customer segment. These tasks will define specifications for a minimum viable design of the robot hardware and software.