# GraSP-VLA：连续场景图驱动VLA长程任务符号规划

> 原文链接：https://robotics.hellobike.com/papers/2025-11-06-2511-04357-5db15f91
> 频道：papers | 更新日期：2025-11-06 | 标签：auto, arxiv, cs.RO
> 核心摘要：针对VLA模型长程规划短板，提出GraSP-VLA神经符号框架。利用连续场景图将人类演示转化为符号表征，推理时自动生成规划域并编排底层VLA策略，实现在真实机器人长程任务中的有效执行与泛化。

# GraSP-VLA：连续场景图驱动VLA长程任务符号规划

::: v-pre
> 本文由 AI 根据公开内容整理，可能有误，请以原文为准：[GraSP-VLA: Graph-based Symbolic Action Representation for Long-Horizon Planning with VLA Policies](https://arxiv.org/abs/2511.04357)（arXiv 2511.04357）

## 核心内容

针对现有VLA模型缺乏高层符号规划导致长程任务能力受限，以及动作模型学习（AML）泛化性不足的问题，arXiv论文提出GraSP-VLA神经符号框架。该方法核心在于引入连续场景图（Continuous Scene Graph）表征，能够从人类演示中自动提取符号化信息，并在推理阶段动态生成新的规划域，作为底层VLA策略的编排器，显著扩展了机器人连续复现动作的数量上限。

技术架构上，GraSP-VLA通过连续场景图弥合了感知与符号规划的语义鸿沟。系统首先从视觉观测中构建场景图并转化为符号表征，据此自动生成适配当前任务的规划域；随后该符号层作为高层控制器，调度低层VLA策略执行具体原子动作。真实世界实验验证了该表征在长程任务编排中的有效性，证明了神经符号方法在解决具身智能长序列决策问题上的工程潜力。

## 关键看点

- **神经符号架构**：提出GraSP-VLA框架，结合连续场景图与VLA，解决纯端到端模型长程规划缺失及纯符号方法泛化差的问题
- **动态规划域生成**：利用连续场景图从人类演示中提取符号表征，支持推理时根据观测自动生成新规划域，无需预定义固定动作集
- **分层策略编排**：符号表征作为高层编排器调度底层VLA策略，突破单步模仿限制，显著提升长程任务中连续动作复现的成功率
- **真机实证有效**：真实机器人实验验证了连续场景图在长程任务编排中的实用性，确认了该表征对提升VLA长序列决策能力的价值

## 工程与产业观察

GraSP-VLA为具身智能长程任务提供了可落地的神经符号范式。通过连续场景图实现“感知-符号-控制”闭环，解决了VLA在复杂时序任务中的规划断层。该思路对构建具备逻辑推理能力的通用操作机器人具有关键参考价值，提示业界在端到端大模型之外，应重视结构化表征对提升系统鲁棒性与可解释性的工程作用。

## 英文摘要 (Abstract)

Deploying autonomous robots that can learn new skills from demonstrations is an important challenge of modern robotics. Existing solutions often apply end-to-end imitation learning with Vision-Language Action (VLA) models or symbolic approaches with Action Model Learning (AML). On the one hand, current VLA models are limited by the lack of high-level symbolic planning, which hinders their abilities in long-horizon tasks. On the other hand, symbolic approaches in AML lack generalization and scalability perspectives. In this paper we present a new neuro-symbolic approach, GraSP-VLA, a framework that uses a Continuous Scene Graph representation to generate a symbolic representation of human demonstrations. This representation is used to generate new planning domains during inference and serves as an orchestrator for low-level VLA policies, scaling up the number of actions that can be reproduced in a row. Our results show that GraSP-VLA is effective for modeling symbolic representations on the task of automatic planning domain generation from observations. In addition, results on real-world experiments show the potential of our Continuous Scene Graph representation to orchestrate low-level VLA policies in long-horizon tasks.

作者：Maëlic Neau, Zoe Falomir, Paulo E. Santos, Anne-Gwenn Bosser, Cédric Buche
:::

## 原始参考来源 (Sources)

- [arXiv 2511.04357：GraSP-VLA: Graph-based Symbolic Action Representation for Long-Horizon Planning with VLA Policies](https://arxiv.org/abs/2511.04357)