ActionGround：冻结VLA策略的训练无关运行时精化
ActionGround: Training-Free Runtime Refinement of Frozen VLA Policies
浏览论文内容
中文总结 AI 辅助
ActionGround提出一种训练无关的神经符号运行时层，通过阶段感知有限状态机和惯性加权欧拉-拉格朗日项精化冻结VLA策略，在多种模型和任务上显著提升成功率、稳定性和鲁棒性。
中文摘要 AI 辅助
视觉-语言-动作（VLA）模型将视觉观察和语言指令直接映射到机器人动作，但它们并未显式表示操作任务的阶段结构或控制执行所需的刚体动力学。我们提出ActionGround，一个神经符号的、训练无关的运行时层，它包裹冻结的VLA策略，无需重训练、微调或权重访问，每个控制步仅增加不到1毫秒的开销。一个符号化的阶段感知有限状态机识别操作阶段（接近、抓取、搬运或放置），并应用特定阶段的基于规则的修正。与此同时，一个始终开启的、惯性加权的欧拉-拉格朗日项将机器人的运动方程纳入每个控制步，而其动力学残差被记录为一致性诊断指标，而非用作门控。我们在OpenVLA、OpenVLA-OFT、Force-VLA和Generalist-VLA上，使用7自由度Franka Panda在十个LIBERO-Spatial拾放任务上评估ActionGround。在跨任务和骨干网络的固定参数下，ActionGround将成功率提升最多6个百分点，稳定性提升最多19.3个百分点，轨迹效率提升最多15%。在单独的Robosuite噪声扫描中，ActionGround在注入动作噪声下提供约10倍的轨迹加加速度鲁棒性提升。在与真实Agilex Piper试验匹配种子的Robosuite仿真对照中，仿真基线成功率从35%提升到95%。物理硬件实验作为定性部署演示呈现；真实机械臂上每次试验的定量成功率留待未来工作。我们的评估仅限于刚体拾放操作。
英文摘要
Vision-Language-Action (VLA) models map visual observations and language instructions directly to robot actions, but they do not explicitly represent the phase structure of manipulation tasks or the rigid-body dynamics governing execution. We present ActionGround, a neuro-symbolic, training-free runtime layer that wraps a frozen VLA policy without retraining, fine-tuning, or weight access, adding less than 1 ms of overhead per control step. A symbolic phase-aware finite-state machine identifies the manipulation phase (approach, grasp, transport, or place) and applies a phase-specific rule-based correction. In parallel, an always-on, inertia-weighted Euler-Lagrange term incorporates the robot's equations of motion into each control step, while its dynamics residual is logged as a consistency diagnostic rather than used as a gate. We evaluate ActionGround across OpenVLA, OpenVLA-OFT, Force-VLA, and Generalist-VLA on ten LIBERO-Spatial pick-and-place tasks using a 7-DoF Franka Panda. With fixed parameters across tasks and backbones, ActionGround improves success rate by up to 6 percentage points and stability by up to 19.3 percentage points, while improving trajectory efficiency by up to 15%. In a separate Robosuite noise sweep, ActionGround provides approximately a 10x improvement in trajectory-jerk robustness under injected action noise. In a matched-seed Robosuite simulation companion to a real Agilex Piper trial, simulated baseline success increases from 35% to 95%. The physical-hardware experiment is presented as a qualitative deployment demonstration; quantitative per-trial success on the real arm is left for future work. Our evaluation is limited to rigid-object pick-and-place manipulation.
