机器人与具身智能领域的最新论文，以及每篇对产业的意义。
Reliable real-time failure detection is critical for deploying learned robotic policies, yet defining failures explicitly is infeasible in open-world settings. This motivates failure detection from nominal demonstrations alone.
Scaling robot manipulation policies requires large amounts of diverse, action-labeled data, yet collecting such data in the real world remains expensive and difficult to scale. Existing methods either depend on human teleoperation, rely on simulation with limited visual realism, or use human videos that lack embodiment-specific annotations.
Automating biological laboratories requires robots to perform dexterous tool interactions under realistic visual conditions while handling dynamic experimental phenomena, presenting significant challenges for robotic manipulation. To enable systematic evaluation under these conditions, we introduce AutoBio+, a simulation benchmark and framework for dexterous manipulation in realistic biological laboratory environments.
World-action models (WAMs) predict a robot's actions together with the observations those actions will produce. However, the strongest models rely on video diffusion transformers that are far too computationally expensive for resource-constrained edge devices.
Existing real-world reinforcement learning methods improve robot policies by combining human corrections with autonomous rollouts. However, accumulated experience varies in its value for policy improvement, and even successful trajectories can contain inefficient or redundant behavior, motivating more selective experience reuse.
This paper presents a dual-quadrotor hook- suspended transport system for cooperative aerial manipulation of a payload that is not rigidly attached to a fixed point along the cable. Two quadrotors are connected by a single inextensible cable that passes through a frictionless hook mounted on the payload, allowing the cable to slide between its two segments.
The NIST Assembly Task Board #1 remains a challenging benchmark for learning-based methods in robotic manipulation. Over short horizons, it demands precise, contact-rich control of small, tight-fitting parts.
Autoregressive robot policies rely on action tokenizers to convert continuous action chunks into compact sequences of discrete prediction targets. Existing tokenizers can reconstruct action chunks from discrete codes, but often leave implicit what each individual token contributes to the reconstructed action.
Collecting dexterous manipulation demonstrations for each robot hand requires repeated task-specific data acquisition. We ask whether one shared corpus of natural human demonstrations can instead serve as a reusable data interface for different dexterous hands, without collecting task demonstrations on each target robot.
Action-conditioned world models offer a promising foundation for generalist robotic intelligence by predicting the consequences of actions. However, existing robotic datasets provide limited coverage of the diverse interaction dynamics required for generalizable world modeling.
World Action Models (WAMs) inherit powerful priors from pretrained video generation models, but their variational autoencoder (VAE) representations are optimized for visual reconstruction rather than manipulation. External visual representations can improve manipulation, but existing methods couple representation choices with different integration mechanisms, obscuring their individual effects.
这是一篇预印本论文，尚未经过同行评审。作者提出两条互补路径：让 GPT-6 Astra 等前沿模型自主生成演示数据并加入纠错片段，训练低延迟本地策略；同时用三层次嵌套语言标注和多方面指令描述训练策略，提升指令跟随能力。在 RoboCasa 365 和 BEHAVIOR-1K 上，层次化与多方面监督组合达到最高分，混合演示与纯人类演示的总体成功率相同，十字填词任务中超时预算内累计动作执行时间提升约 3.4 倍。
这篇预印本提出 AdaTempo，一种自监督方法，通过提取多个演示中共享的相对节奏模式，对演示轨迹进行离线重采样，训练出更快执行的视觉运动策略。在5个仿真和6个真机任务上，AdaTempo 最高实现3.57倍加速，仿真平均成功率达75.5%、平均加速2.83倍，真机汇总成功率从58.3%提升到77.8%且平均快2.02倍。该方法无需运行时速度选择或在线重定时，简化部署。
这篇预印本提出Proprioceptive Action Models (PAM)，用弧长参数化的B样条草图表示机器人剩余关节路径，并与短动作块联合去噪生成。在Push-T、LIBERO-Long仿真和四个真机双臂任务中，PAM均优于纯动作基线；真机平均成功率从47.5%提高到75.0%。该方法为长程操作提供了紧凑且免时序的运动意图表示，但尚未经过同行评审。
本文是预印本，提出PEWAM，一个71.5M参数的世界行动模型。它把单条演示编译成不含机器人本体的效应程序，包括物体关键点轨迹、两个接触点和终态，并在闭环中从现场重求解执行。在LIBERO-Goal留出任务上，单条演示程序完成40/90回合，而同日标的基线最多完成19；该程序可在四个仿真臂和真实Franka上跨本体运行，并能吸收5厘米推动。
本文为预印本。DynaHarness通过动态物理执行合同连接慢速语义推理和快速物理治理，在LIBERO-Pro新增800个状态上取得75.2%成功率，冻结π0.5策略仅17.5%。同一技能库下，完整动态执行74.0%，对比一次性重规划63.9%，显示对已有能力进行实时监控、替换和恢复的价值。
Dream4ACT 提出 action views，把目标关节构型渲染为四路固定虚拟相机图像，让不同关节空间机器人共享同一视频生成接口，并通过无训练的 URDF 约束匹配还原可执行关节目标。模型在 RoboTwin2.0 仿真平均成功率为 88.98%，TriWorldBench 总分 65.66，真机四平台两项常见任务成功率分别为 87.5%、85.0%、87.5%、90.0%。本文为预印本，尚未经过同行评审。
RoboCoach 是一个预印本框架，用世界模型中的想象执行来诊断长程机器人任务中最先失败的子任务，并据此请求对应技能专家的演示数据。在 Franka 和 AgileX 真机上，仅追加 150 条子任务演示，成功率分别从 13.3% 升至 75.0%、从 40.0% 升至 83.8%，显著超过统一采集加共享适配器的基线。该工作表明世界模型可以成为主动教练，把稀缺的真实数据导向可复用的技能模块。
World-Action Models (WAMs) couple visual dynamics with action prediction, bringing the rich priors of pretrained video models to robotic manipulation. However, their multi-view interfaces typically tile images or concatenate tokens, leaving the geometric relationships among synchronized cameras implicit.
Robot policies are frequently trained from human corrections, yet teleoperating a robot to provide corrections is burdensome, and human demonstrators are not always optimal. We propose Blended DAgger (BlenDAgger), an approach for collecting data to train imitation learning policies by using shared control to blend the policy's and demonstrator's actions during interventions.
论文来自 arXiv 机器人领域。尚未生成摘要的论文，显示摘要开头部分。