Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.23620 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913063032061952 |
|---|---|
| author | Xu, Haoming Lei, Lei Gu, Jie Tang, Chu Chen, Jingmin Wang, Ruiqi |
| author_facet | Xu, Haoming Lei, Lei Gu, Jie Tang, Chu Chen, Jingmin Wang, Ruiqi |
| contents | We present Move-Then-Operate, a Vision language action framework that explicitly decouples robotic manipulation into two distinct behavioral phases: coarse relocation (move) and contact-critical interaction (operate). Unlike monolithic policies that conflate these heterogeneous regimes, our architecture employs a dual-expert policy routed by a learnable phase selector, introducing a structural inductive bias that isolates phase-specific dynamics. Phase labels are automatically generated via an MLLM-based pipeline conditioned on lightweight contextual cues such as end-effector velocity and subtask decomposition to ensure alignment with human motor patterns. Evaluated on the RoboTwin2 benchmark, our method achieves an average success rate of $68.9\%$, outperforming the monolithic $π_0$ baseline by $24\%$. It matches or exceeds models trained on $10\times$ more data and reaches peak performance in $40\%$ fewer training steps, demonstrating that architectural disentanglement of move and operate phases is a highly effective and efficient strategy for mastering high-precision manipulation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_23620 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Move-Then-Operate: Behavioral Phasing for Human-Like Robotic Manipulation Xu, Haoming Lei, Lei Gu, Jie Tang, Chu Chen, Jingmin Wang, Ruiqi Robotics We present Move-Then-Operate, a Vision language action framework that explicitly decouples robotic manipulation into two distinct behavioral phases: coarse relocation (move) and contact-critical interaction (operate). Unlike monolithic policies that conflate these heterogeneous regimes, our architecture employs a dual-expert policy routed by a learnable phase selector, introducing a structural inductive bias that isolates phase-specific dynamics. Phase labels are automatically generated via an MLLM-based pipeline conditioned on lightweight contextual cues such as end-effector velocity and subtask decomposition to ensure alignment with human motor patterns. Evaluated on the RoboTwin2 benchmark, our method achieves an average success rate of $68.9\%$, outperforming the monolithic $π_0$ baseline by $24\%$. It matches or exceeds models trained on $10\times$ more data and reaches peak performance in $40\%$ fewer training steps, demonstrating that architectural disentanglement of move and operate phases is a highly effective and efficient strategy for mastering high-precision manipulation. |
| title | Move-Then-Operate: Behavioral Phasing for Human-Like Robotic Manipulation |
| topic | Robotics |
| url | https://arxiv.org/abs/2604.23620 |