Robotics · 2026-10-22
CW-WAM: A Structured Contact-Wrench World Action Model for Force-Sensitive Manipulation
Houzhu Ding, Xinda Qi, Luqi Wang, Ming Yang
AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.
TL;DR
CW-WAM is a world action model for force-sensitive manipulation that jointly predicts video, motion, and contact wrench. It uses a unified interaction module with paired action–wrench history, pair-causal attention, and long-term wrench memory, and drives a 20-Hz outer-loop admittance controller in a slow–fast system. On real robots, it raises insertion success from 25% to 70% (peg) and from 40% to 95% (adapter) vs WAM without wrench input.
Why it matters
The authors address limits of visual-only action models for contact-rich tasks, where load and contact state affect outcomes. CW-WAM predicts an expected wrench trajectory and uses it directly as a reference for admittance feedback, enabling smoother force tracking and task success improvements across multiple real-robot contact tasks, plus better lift-height accuracy under different payloads.
Method
- Jointly predicts future video latents, motion targets, and contact wrench; training pairs each motion target with a compensated wrench at the same timestamp.
- Interaction module: paired motion–wrench history, pair-causal attention (future wrench tokens attend to paired past motion–wrench tokens), and long-term wrench memory (LTM) for earlier contact information.
- Slow–fast execution: low-rate CW-WAM inference produces nominal pose and a selected predicted wrench chunk; a 20-Hz outer-loop admittance controller applies corrections while a 1-kHz joint controller tracks interpolated trajectories.
Limitation
The authors state that their evaluation uses manually tuned feedback gains and predefined compliant directions and that future work will examine how controller settings affect performance.
Abstract (from arXiv)
Force-sensitive manipulation requires contact and load information that is not provided explicitly by visual observations. We present CW-WAM, a world action model that jointly predicts video, motion, and contact wrench. Inspired by hybrid motion-force control, it contains a unified interaction module that treats both motion and contact wrench as modeling targets and predicts time-aligned motion-wrench chunks as control references. A pair-causal attention mechanism structures the motion-wrench dependencies, while a long fixed-length wrench memory retains past interaction information. The predicted wrench chunk acts as a force reference for a 20-Hz outer-loop admittance controller within a slow-fast control system that maintains force feedback and smooth motion between model queries. CW-WAM improves contact prediction and increases the insertion success rate from 25% to 70% for square pegs with 1-mm clearance per side and from 40% to 95% for power adapters, compared with WAM without wrench input. It also improves lifting-height accuracy across different loads. Our framework with an admittance controller reduces force-reference RMSE from 4.8 to 2.7 N for peg insertion and from 2.0 to 1.3 N for wiping, respectively, with smaller force fluctuations during successful execution. Project website: https://cw-wam-project.github.io/
Related papers
- ForceWAM: Decoupling Future and Response Learning from Demo and Play Data for Contact-Rich Manipulation
- ForceLDM: A Force-aware Latent Dynamics Model for Contact-Rich Manipulation
- Dexterous Tactile World Model
- ACTS: Learning Visuo-Tactile World Models from Human Physical Interaction
- UMI-WM: Interactive World Modeling for Bimanual Manipulation from Wrist Cameras
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First public release: 2026-10-21 · Indexed by Pub-AI: 2026-10-22 UTC
CW-WAM: A Structured Contact-Wrench World Action Model for Force-Sensitive Manipulation
Preprint posted: 2026-10-21 · Original record ↗
- arxiv
- 2610.30117
Explicit source versions (1)
- v1 · 2026-10-21
- Preprint available
Source date: 2026-10-21 · First detected: 2026-10-25 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · atom:published (official metadata)
Abstract · arxiv
Force-sensitive manipulation requires contact and load information that is not provided explicitly by visual observations. We present CW-WAM, a world action model that jointly predicts video, motion, and contact wrench. Inspired by hybrid motion-force control, it contains a unified interaction module that treats both motion and contact wrench as modeling targets and predicts time-aligned motion-wrench chunks as control references. A pair-causal attention mechanism structures the motion-wrench dependencies, while a long fixed-length wrench memory retains past interaction information. The predicted wrench chunk acts as a force reference for a 20-Hz outer-loop admittance controller within a slow-fast control system that maintains force feedback and smooth motion between model queries. CW-WAM improves contact prediction and increases the insertion success rate from 25% to 70% for square pegs with 1-mm clearance per side and from 40% to 95% for power adapters, compared with WAM without wrench input. It also improves lifting-height accuracy across different loads. Our framework with an admittance controller reduces force-reference RMSE from 4.8 to 2.7 N for peg insertion and from 2.0 to 1.3 N for wiping, respectively, with smaller force fluctuations during successful execution. Project website: https://cw-wam-project.github.io/
