arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新，休息一下吧，周末愉快~~
arXiv 2602.02459cs.RO
收藏

TIC-VLA：一种用于动态环境中机器人导航的思考控制视觉-语言-动作模型

TIC-VLA: A Think-in-Control Vision-Language-Action Model for Robot Navigation in Dynamic Environments

University of California, Los Angeles（加州大学洛杉矶分校）

Zhiyu Huang, Yun Zhang, Johnson Liu, Rui Song, Chen Tang, Jiaqi Ma

阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出TIC-VLA模型，通过显式建模推理延迟并引入延迟语义-控制接口，结合异步训练流程，解决动态环境中视觉-语言-动作模型的推理与实时控制异步问题，在仿真和真实机器人上优于先前模型。

摘要：

在动态、以人为中心的环境中，机器人必须遵循语言指令同时保持实时反应控制。视觉-语言-动作（VLA）模型提供了一个有前景的框架，但它们假设时间对齐的推理和控制，尽管语义推理相对于实时动作固有地存在延迟。我们提出了Think-in-Control（TIC）-VLA，一种延迟感知框架，在动作生成过程中显式建模延迟的语义推理。TIC-VLA定义了一个延迟语义-控制接口，该接口除了当前观测外，还基于延迟的视觉-语言语义状态和显式延迟元数据来条件化动作生成，使策略能够补偿异步推理。我们进一步提出了一种延迟一致的训练流程，在模仿学习和在线强化学习期间注入推理推理延迟，使训练与异步部署对齐。为了支持现实评估，我们提出了DynaNav，一个物理精确、照片级真实的仿真套件，用于动态环境中的语言引导导航。在仿真和真实机器人上的大量实验表明，TIC-VLA在多秒推理延迟下始终优于先前的VLA模型，同时保持鲁棒的实时控制。项目网站：https://ucla-mobility.github.io/TIC-VLA/

英文摘要：

Robots in dynamic, human-centric environments must follow language instructions while maintaining real-time reactive control. Vision-language-action (VLA) models offer a promising framework, but they assume temporally aligned reasoning and control, despite semantic inference being inherently delayed relative to real-time action. We introduce Think-in-Control (TIC)-VLA, a latency-aware framework that explicitly models delayed semantic reasoning during action generation. TIC-VLA defines a delayed semantic-control interface that conditions action generation on delayed vision-language semantic states and explicit latency metadata, in addition to current observations, enabling policies to compensate for asynchronous reasoning. We further propose a latency-consistent training pipeline that injects reasoning inference delays during imitation learning and online reinforcement learning, aligning training with asynchronous deployment. To support realistic evaluation, we present DynaNav, a physics-accurate, photo-realistic simulation suite for language-guided navigation in dynamic environments. Extensive experiments in simulation and on a real robot show that TIC-VLA consistently outperforms prior VLA models while maintaining robust real-time control under multi-second reasoning latency. Project website: https://ucla-mobility.github.io/TIC-VLA/

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑