Hierarchical policies that combine language and low-level control have been shown to perform impressively long-horizon robotic tasks, by leveraging either zero-shot high-level planners like pretrained language and vision-language models (LLMs/VLMs) or models trained on annotated robotic demonstrations. However, for complex and dexterous skills, attaining high success rates on long-horizon tasks still represents a major challenge -- the longer the task is, the more likely it is that some stage will fail. Can humans help the robot to continuously improve its long-horizon task performance through intuitive and natural feedback? In this paper, we make the following observation: high-level policies that index into sufficiently rich and expressive low-level language-conditioned skills can be readily supervised with human feedback in the form of language corrections. We show that even fine-grained corrections, such as small movements ("move a bit to the left"), can be effectively incorporated into high-level policies, and that such corrections can be readily obtained from humans observing the robot and making occasional suggestions. This framework enables robots not only to rapidly adapt to real-time language feedback, but also incorporate this feedback into an iterative training scheme that improves the high-level policy's ability to correct errors in both low-level execution and high-level decision-making purely from verbal feedback. Our evaluation on real hardware shows that this leads to significant performance improvement in long-horizon, dexterous manipulation tasks without the need for any additional teleoperation. Videos and code are available at https://yay-robot.github.io/.
Naval Research Laboratory（海军研究实验室）；University of Wisconsin–Madison（威斯康星大学麦迪逊分校）；University of Washington（华盛顿大学）；University of Wisconsin–La Crosse（威斯康星大学拉克罗斯分校）
Learning to Fly via Deep Model-Based Reinforcement Learning
Volkswagen Group, Machine Learning Research Lab（大众集团机器学习研究实验室）；Technical University of Darmstadt（达姆施塔特工业大学）；Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）
On-the-fly hand-eye calibration for the da Vinci surgical robot
Department of Mechanical Engineering, Imperial College London（帝国理工学院机械工程系）；Mechatronics in Medicine Laboratory（医学机电实验室）；Hamlyn Centre for Robotics Surgery（机器人外科哈姆林中心）
Why Evolve When You Can Adapt? Post-Evolution Adaptation of Genetic Memory for On-the-Fly Control
School of Engineering and Physical Sciences, Heriot-Watt University Dubai（工程与物理科学学院，惠顿大学迪拜分校）；School of Mathematical and Computer Sciences, Heriot-Watt University Dubai（数学与计算机科学学院，惠顿大学迪拜分校）
On-the-Fly VLA Adaptation via Test-Time Reinforcement Learning
University of Missouri–Kansas City（密苏里大学堪萨斯城分校）；Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；U. S. Naval Research Laboratory（美国海军研究实验室）；Lamar University（拉马尔大学）；Meta AI；Rochester Institute of Technology（罗彻斯特理工学院）
Fly-by-Code: Embodied Coding Agents for Aerial Manipulation with Active Visual and Physical Feedback
Seoul National University（首尔国立大学）；Carnegie Mellon University（卡内基梅隆大学）；University of Maryland, College Park（马里兰大学帕克分校）；All Purpose AI；Kyung Hee University（庆熙大学）
Deployment of Aerial Robots after a major fire of an industrial hall with hazardous substances, a report
University of Applied Science, Gelsenkirchen（盖尔森基兴应用科学大学）；German Rescue Robotic Centre, Dortmund（多特蒙德德国救援机器人中心）；Fire Department of Dortmund, Institute of Fire Service and Rescue Technology（多特蒙德消防局消防与救援技术研究所）
