rMuscle: Robotic Muscle Memory for Efficient Vision-Language-Action Model Inference

工厂重复性劳动是具身智能早期落地的重要场景，结构化工位有利于当前策略执行任务，而 VLA 推理延迟直接影响机器人响应和运动连续性。现有框架没有充分利用具身工作负载特征，也没有分别处理不同推理阶段的瓶颈。作者发现，重复执行任务时，相似性不仅存在于观察与动作轨迹中，也存在于模型内部状态中。据此提出受肌肉记忆启发的实时推理框架 rMuscle：上下文缓存复用视觉 token 输出以减少计算，动作缓存复用神经元激活模式以减少权重访问，并通过在线缓存重建、滑动窗口检索和连续去噪步共享掩码控制内存与访问开销。摘要报告在 RTX 4090 和 Jetson Thor、LIBERO、RoboTwin 与真实操作任务上达到 1.29–1.42 倍加速，并保持真实机器人原始成功率。注意此区间概括核心配置，正文另给出不同模型和基线下的结果。

Factory work is a promising early scenario for embodied AI: assigning repetitive manual jobs to robots has clear economic payoff, and a structured station keeps the jobs tractable for current policies. Vision-Language-Action (VLA) models now dominate as the policy paradigm for these robots. The inference latency of VLA models directly affects robot responsiveness and motion smoothness. However, existing VLA inference frameworks do not fully exploit the characteristics of embodied workloads or account for the distinct bottlenecks across different stages of VLA inference. In this paper, we first characterize embodied workloads and identify substantial task similarity across repeated robot executions. We further find that such similarity extends beyond observations and action trajectories to internal model states. Drawing on these observations, we present rMuscle, a real-time VLA inference framework inspired by human muscle memory. It exploits cross-execution similarity through a dual-phase muscle-memory cache. The Context Cache reuses visual-token outputs to reduce computation, while the Action Cache reuses neuron activation patterns to reduce weight accesses. We keep both the cache memory footprint and access overhead low through online cache recomputation, sliding-window cache retrieval, and mask sharing across consecutive denoising steps. rMuscle achieves 1.29–1.42× speedup on RTX 4090 and Jetson Thor across LIBERO, RoboTwin, and physical manipulation tasks, while maintaining the original success rates on real-world robots.