让大模型机器人动作实时反应，40毫秒刷新一次感知
$π\mathbf{R}^2$: Reactive Real-time Flow Policies
- 用快速本体感觉+慢速视觉分路，实现动作中即时响应
- 单步去噪生成动作，硬件延迟自适应，实测达25Hz
- 无需改架构，真实机械臂成功率达基线30%以上
通用操作策略正采用基于大型预训练模型的动作块流水线结构。但这些动作块以开环运行，无法在执行中响应感官输入，丧失了反应能力。频繁重规划可恢复反应性，但感知到动作的延迟管道（大型骨干网络加多步去噪）过慢，阻碍高频重规划，导致已执行动作过时，不适用于动态闭环控制。我们提出 $π\mathbf{R}^2$，在保留大型骨干、多模态表达和多动作预测能力的同时，使策略具备反应性和实时性。基于扩散强制的逐位置噪声调度，$π\mathbf{R}^2$ 提出两项创新：首先，将条件输入分为快速通道（每周期更新的本体感觉）与异步更新的慢速通道（视觉-语言特征），使策略在动作块内响应本体感觉，同时容忍过时的视觉信息；其次，采用延迟自适应的流动调度，将飞行中的动作视为补全条件，每次调用仅需一步去噪即可输出动作，使单一模型能适配不同硬件延迟。仅需对现有架构做极小修改，$π\mathbf{R}^2$ 可从预训练策略微调得到。在真实 xArm6+XHand 平台上应用 GR00T-N1.7 模型，其闭环重规划速度比基线快约4倍（约25Hz，A5000 GPU），每40ms处理一次新观测。在仿真和真实世界操作任务中，$π\mathbf{R}^2$ 的成功率相比最强基线提升最高达23%（仿真）和30%（真实世界）。
原文摘要 · Abstract (English)
Generalist manipulation policies increasingly take the form of action-chunking flow policies built on large pretrained backbones. Such chunks run open-loop, so the policy cannot react to sensory input arriving mid-execution, sacrificing \emph{reactivity}. Replanning more often would restore it, but the perception-to-action pipeline (a large backbone plus multiple denoising steps) is too slow: this \emph{latency} forbids frequent replanning and leaves committed actions stale, making such policies ill-suited for dynamic, closed-loop control. We present $π\mathbf{R}^2$, which makes these policies reactive and real-time while retaining large backbones, expressive multi-modal policies, and multi-action prediction. Built on the per-position noise schedule of diffusion forcing, $π\mathbf{R}^2$ contributes two ideas. First, it splits conditioning into a fast channel (proprioception, fresh every tick) and an asynchronously updated slow channel (vision-language features), so the policy reacts to proprioception within a chunk while tolerating stale vision. Second, a latency-adaptive flow schedule treats in-flight actions as inpainting conditioning and emits actions in one denoising step per call, letting one trained model adapt to varying hardware latency. Requiring minimal modification to existing architectures, $π\mathbf{R}^2$ can be finetuned from a pretrained policy: applied to GR00T-N1.7 on a real xArm6+XHand platform, it replans closed-loop roughly $4\times$ faster than the base policy (~$25$Hz on an A5000 GPU), acting on a fresh observation every $40$ms. Across simulation and real-world manipulation tasks, $π\mathbf{R}^2$ improves the success rate by up to $23\%$ in simulation and $30\%$ in the real world over the strongest baseline. Project page: https://pi-r2-flow.github.io/
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。