ForceVLA2：释放面向接触丰富操作的力觉感知混合力位控制
ForceVLA2: Unleashing Hybrid Force-Position Control with Force Awareness for Contact-Rich Manipulation
把力信号作为提示注入 VLM 专家、并用跨尺度 MoE 与实时力做闭环融合，为视觉-语言-动作模型引入力觉通道提供了可复用的架构范式；其多视角图像+力信号的数据集也可用于研究接触丰富场景下的多模态表征。
原文依据 employs a Cross-Scale Mixture-of-Experts (MoE) in the action expert to adaptively fuse these concepts with real-time interaction forces
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论面向接触丰富操作的具身智能此前主要依赖位置控制，而对交互力的显式感知与调节仍研究不足，限制了真实任务中的稳定性、精度与鲁棒性。我们提出 ForceVLA2，一个端到端的视觉-语言-动作框架，使机器人具备混合力位控制与显式力觉感知能力。ForceVLA2 将基于力的提示引入 VLM 专家，以跨阶段构建力觉感知的任务概念，并在动作专家中采用跨尺度混合专家（MoE），将这些概念与实时交互力自适应融合，实现闭环的混合力位调节。为支撑学习与评测，我们构建了 ForceVLA2-Dataset，包含 5 个接触丰富任务（包括擦拭、按压与装配）上的 1000 条轨迹，并带有多个视角的图像、任务提示、本体感受状态与力信号。大量实验表明，ForceVLA2 显著提升了接触丰富操作的成功率与可靠性，在 5 个任务上分别比 pi0 和 pi0.5 高出 48.0% 和 35.0%，并缓解了机械臂过载与接触不稳定等常见失效模式，从而有力推动了 VLA 中力觉感知的交互式物理智能发展。
英文原文摘要 1323 字符
Embodied intelligence for contact-rich manipulation has predominantly relied on position control, while explicit awareness and regulation of interaction forces remain under-explored, limiting stability, precision, and robustness in real-world tasks. We propose ForceVLA2, an end-to-end vision-language-action framework that equips robots with hybrid force-position control and explicit force awareness. ForceVLA2 introduces force-based prompts into the VLM expert to construct force-aware task concepts across stages, and employs a Cross-Scale Mixture-of-Experts (MoE) in the action expert to adaptively fuse these concepts with real-time interaction forces for closed-loop hybrid force-position regulation. To support learning and evaluation, we construct ForceVLA2-Dataset, containing 1,000 trajectories over 5 contact-rich tasks, including wiping, pressing, and assembling, with multi-view images, task prompts, proprioceptive state, and force signals. Extensive experiments show that ForceVLA2 substantially improves success rates and reliability in contact-rich manipulation, outperforming pi0 and pi0.5 by 48.0% and 35.0%, respectively, across the 5 tasks, and mitigating common failure modes such as arm overload and unstable contact, thereby actively advancing force-aware interactive physical intelligence in VLAs.