VLA 模型全生命周期综述发布:覆盖 88 篇工作、分析 22 个模型
Vision-Language-Action Models: A Life-Cycle Survey of Design, Training and Deployment
AI 导读
arXiv 论文 2611.18372 发布 VLA 模型生命周期综述,回顾 2022 年 12 月至 2026 年 10 月的 88 篇工作,沿七个轴分析 22 个模型的设计、训练与部署。
正文 · 原文
Abstract:Vision-language-action (VLA) models adapt web-pretrained vision-language models to output robot actions, so that web knowledge carries over to physical control. In under 4 years the field has moved from RT-2, the first VLA, to open models that fine-tune on a single consumer GPU and to proprietary generalists that clean unseen homes. This survey reviews 88 works from December 2022 to October 2026, mainly on manipulation and navigation, plus humanoid control, world-action models and evaluation. We follow a VLA through its life cycle, from model design to training and deployment, along seven axes. Model design covers (a) inputs and memory, what the model observes; (b) architecture, how its pipeline is built; (c) action output, how it produces motor commands; and (d) intermediate predictions, what else it predicts, from reasoning steps to future observations. Training covers (e) pretraining data, the mix of robot, web, human and simulated data, and (f) post-training, adaptation to the target robot. Deployment covers (g) real-time execution, keeping pace with the robot. We analyse 22 models on these axes and report their training data, model size, inference speed and hardware. Three trends emerge: (i) training data has broadened, from about 10,000 hours of teleoperated robot data ($\pi_0$) to 23,000 community-collected episodes (SmolVLA) and 20,000 hours of egocentric human video (GR00T N1.7); (ii) model sizes have split: most recent models have 3-7 billion parameters, while a new group under one billion runs on consumer GPUs; and (iii) generalist VLAs are emerging that handle both manipulation and navigation, such as UniVLA and Qwen-VLA. We also compare VLAs with world-action models, which predict future observations as well as actions. We review evidence that some VLAs with near-perfect benchmark scores fail when objects or the camera move slightly, and open up further research questions.
| Comments: | 43 pages, 9 figures, 11 tables |
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Machine Learning (cs.LG) |
| Cite as: | arXiv:2611.18372 [cs.RO] |
| (or arXiv:2611.18372v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2611.18372 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Suvarna Kadam [view email]
[v1]
Fri, 13 Nov 2026 16:17:00 UTC (91 KB)
来源:arXiv cs.AI · arxiv.org