# 从滑动窗口到自演化终身记忆：长时程机器人 VLA 模型记忆范式的双视角综述

> 原文链接：https://robotics.hellobike.com/papers/2026-10-16-2610-23319
> 频道：papers | 更新日期：2026-10-16 | 标签：arxiv, cs.RO
> 核心摘要：该综述针对 VLA 长时程任务记忆不足，按范围与持久性建立分类，并在代表性系统上比较分析。

# 从滑动窗口到自演化终身记忆：长时程机器人 VLA 模型记忆范式的双视角综述

::: v-pre
> 原文链接：[From Sliding Windows to Self-evolving Lifelong Memory: A Dual-perspective Survey of Memory Paradigms for Long-Horizon Robotics VLA Models](https://arxiv.org/abs/2610.23319)（arXiv 2610.23319 · 发布日期：2026-10-16）

**From Sliding Windows to Self-evolving Lifelong Memory: A Dual-perspective Survey of Memory Paradigms for Long-Horizon Robotics VLA Models**

该综述针对 VLA 长时程任务记忆不足，按范围与持久性建立分类，并在代表性系统上比较分析。

## 核心要点

- 长时程机器人任务暴露 VLA 时间上下文受限，难以积累并复用历史经验。
- 按记忆范围与持久性构建系统分类，归纳五类代表性范式和一个新兴前沿。
- 比较记忆特性、架构复杂度与开源成熟度，指出复杂度升高通常伴随开源成熟度下降。
- 讨论全生命周期记忆系统的关键挑战，包括高效记忆管理与终身知识演化。

**验证范围与局限**：范围限于对代表性 VLA 记忆系统的综述比较。

## 英文摘要 (Abstract)

Long-horizon robotic tasks expose a fundamental limitation of current Vision-Language-Action (VLA) models because their restricted temporal context pre-vents effective accumulation and reuse of past experiences. As a result, memory has become a fundamental component for enabling embodied agents to perform long-horizon reasoning and adaptive decision-making. This survey presents a systematic taxonomy of VLA memory mechanisms based on memory scope and persistence, categorizing existing approaches into five representative paradigms and one emerging frontier. Based on this taxonomy, we compare representative VLA memory systems in terms of memory characteristics, architectural com-plexity, and open-source maturity, revealing a clear trade-off in which increasing architectural complexity is generally accompanied by decreasing open-source ma-turity. Finally, we discuss the key challenges in building full-lifecycle memory systems, including efficient memory management and lifelong knowledge evolu-tion. Overall, this survey provides a unified perspective on memory architectures for VLA agents and offers insights into the development of lifelong embodied in-telligence.

作者：Yong Wang, Boyu Shen, Xinwei Zhang, Yizheng Li, 罗科元
:::

## 原始参考来源 (Sources)

- [arXiv 2610.23319：From Sliding Windows to Self-evolving Lifelong Memory: A Dual-perspective Survey of Memory Paradigms for Long-Horizon Robotics VLA Models](https://arxiv.org/abs/2610.23319)