DECOWAM: Decoupled Whole-Body World-Action Model for Legged Mobile Manipulation

DECOWAM: Decoupled Whole-Body World-Action Model for Legged Mobile Manipulation

|Authors: Siyuan Ma, Boshi Zhang, Yutian Zhang, Qinglian Wu, Jiaqi Zhai, Dong Wei, Qiaojun Yu

|Venue: arXiv preprint | Year: 2026

|arXiv: 2608.20114v1

Abstract

Mobile manipulation requires a robot to predict how locomotion and arm motion jointly alter future observations and control. Existing world-action models, developed largely for fixed-base platforms, do not explicitly distinguish camera ego-motion from base and arm actions. Here we introduce DECOWAM, a whole-body world-action model that separates these factors through dedicated conditional interfaces. DECOWAM freezes an adapted FastWAM backbone and trains residual adapters, an action-equivalent future bottleneck distilled from privileged observations, adversarially separated base and arm latents, and base-velocity conditioning for video prediction. We further introduce ARMDOG, a real-robot dataset that synchronizes video, whole-body state and action, and language. On a fixed replay protocol, DECOWAM improved both future-video and action prediction over FastWAM, reducing action MSE by 21.7% with 25.95M trainable adaptation parameters. Across 79 closed-loop trials per method, it achieved the highest observed whole-body coordination and base-displacement robustness among the compared systems, while task completion remained comparable to the strongest baseline. These results show that embodiment-aware factorization can support parameter-efficient joint visual prediction and whole-body control under moving viewpoints.

Key Contributions

• Mobile manipulation requires a robot to predict how locomotion and arm motion jointly alter futur…
• Existing world-action models, developed largely for fixed-base platforms, do not explicitly disti…
• Here we introduce DECOWAM, a whole-body world-action model that separates these factors through d…
• DECOWAM freezes an adapted FastWAM backbone and trains residual adapters, an action-equivalent fu…
• We further introduce ARMDOG, a real-robot dataset that synchronizes video, whole-body state and a…

Topics

• manipulation
• locomotion
• vision
• reinforcement-learning
• control
• benchmark

Code & Data

No code repository linked in paper metadata.

BibTeX

@article{Ma2026_260820114v1,
  title     = {DECOWAM: Decoupled Whole-Body World-Action Model for Legged Mobile Manipulation},
  author    = {Siyuan Ma and Boshi Zhang and Yutian Zhang and Qinglian Wu and Jiaqi Zhai and Dong Wei and Qiaojun Yu},
  year      = {2026},
  eprint    = {2608.20114v1},
  archivePrefix = {arXiv},
  primaryClass  = {cs.AI},
  url       = {https://arxiv.org/abs/2608.20114v1}
}