This paper explores the challenges faced by assistive robots in effectively cooperating with humans, requiring them to anticipate human behavior, predict their actions' impact, and generate understandable robot actions. The study focuses on a use-case involving a user with limited mobility needing assistance with pouring a beverage, where tasks like unscrewing a cap or reaching for objects demand coordinated support from the robot. Yet, anticipating the robot's intentions can be challenging for the user, which can hinder effective collaboration. To address this issue, we propose an innovative solution that utilizes Augmented Reality (AR) to communicate the robot's intentions and expected movements to the user, fostering a seamless and intuitive interaction.

翻译：本文探讨了辅助机器人在与人类有效协作时面临的挑战，要求机器人能够预测人类行为、预判自身动作的影响，并生成可被理解的行动。研究聚焦于一个典型案例：行动受限的用户需要机器人协助倾倒饮料，其中拧开瓶盖或取物等任务需要机器人提供协调支持。然而，用户难以预判机器人的意图，这阻碍了有效协作。为解决此问题，我们提出一种创新方案，利用增强现实（Augmented Reality, AR）向用户传达机器人的意图与预期动作，从而促进流畅且直观的交互。