具身开发者社区

外观

论文arXiv 2610.1148020 天前

RoboAware：从反事实结果中学习协调具身技能 ​

RoboAware针对具身技能组合中的策略族选择，借反事实分支结果学习状态条件协调器，并在100项单回合任务上验证。

AI 整理
  • 数据与采集
  • 具身模型
  • 端侧部署

原文链接：RoboAware: Learning to Coordinate Embodied Skills from Counterfactual Outcomes（arXiv 2610.11480 · 发布日期：2026-10-08）

RoboAware: Learning to Coordinate Embodied Skills from Counterfactual Outcomes

RoboAware针对具身技能组合中的策略族选择，借反事实分支结果学习状态条件协调器，并在100项单回合任务上验证。

核心要点 ​

  • 问题：模块化机器人技能可与冻结端到端策略组合，但需预判当前状态下哪类策略会成功。
  • 方法：提出P^5模式与层级MDP，并用SCB在同一训练状态下生成各可用策略族的反事实结果。
  • 学习：EAL结合蒙特卡洛树搜索与Q学习，将分支结果蒸馏为策略族条件化价值。
  • 结果：在100项单回合评估中，总成功率77.0%，并优于code-as-policy和VLA-harness基线。

验证范围与局限：验证范围为100项单回合任务，涵盖RoboSuite、LIBERO-Pro任务簇和RoboTwin双臂任务。

英文摘要 (Abstract) ​

Embodied coding agents can combine modular robot skills with frozen end-to-end policies, yet effective composition requires anticipating which policy family will succeed in the current physical state. We present RoboAware, which builds on coding agents' skill orchestration by learning only a state-conditioned responsibility coordinator from counterfactual outcomes. Inspired by the success of REPL, we propose the $P^5$ schema and formulate a hierarchical MDP based on it. $P^5$ organizes skills uniformly into five semantic stages, defining where responsibility can be compared. To address the lack of counterfactual branch outcomes in existing work, we introduce State-Locked Counterfactual Branching (SCB), which restores the same training state to generate and execute a code block from each admissible family, exposing outcomes that selected-branch experience leaves unobserved. Building on this, we propose Execution-Aware Learning (EAL), which combines Monte Carlo tree search with Q-learning to distill these outcomes into family-conditioned values. At deployment, the coordinator selects the policy family according to observable context, and the frozen coding agent generates the next local code block. Comprehensive single-episode evaluations on 100 tasks show that RoboAware reaches a 77.0% overall success rate, with SOTA averages of 90.0% on RoboSuite, 73.8% on diverse LIBERO-Pro task clusters, and 90.0% on challenging RoboTwin bimanual tasks, outperforming existing code-as-policy and VLA-harness baselines.

作者：Bohan Zhou, Xingbei Chen, Emily Huang, Weilin Ruan, Haojian Huang, Y. Zhang, Zexi Li, Wenqian Li, Qize Yu, Zetian Song

来源

本页事实对应的公开页面，以原文为准。

  1. arXiv 2610.11480：RoboAware: Learning to Coordinate Embodied Skills from Counterfactual Outcomes访问 2026-10-21