AI 中文总结
研究针对长期机器人规划中对具身经验记忆的需求，提出MEMORA方法，通过形成-巩固-检索生命周期及四种类型存储实现具身动作记忆，经实验评估取得良好结果，能为机器人规划提供记忆上下文。
AI 中文摘要
长期的机器人规划不仅需要预测下一步行动，还需要对具身经验的记忆以使未来目标可解释。人们并非仅从当前场景进行规划，还会借助记忆中的地点、物体状态变化、先前程序及反复行动揭示的规律。我们将具身动作记忆（EAM）定义为形成、维持并使用此类经验作为持久记忆状态以供后续决策的能力。MEMORA通过形成-巩固-检索生命周期及四种类型的存储来实现EAM，包括环境记忆、实体记忆、活动记忆和推理知识。在线编辑在新观察到来时维护物体身份和状态历史，离线巩固将反复经验抽象为可重用程序和特定参与者规律。MEMORA-Bench通过基于记忆的规划（包括未见目标）和补充记忆评估任务，在45小时的EPIC-KITCHENS-100扩展视频上对18名参与者评估此生命周期。在四个开放权重语言模型中，完整的MEMORA（结合编辑、类型存储和巩固）在评估的记忆条件中取得最强综合结果。它将记忆评估准确率比最强控制基线提高多达20.5分，将分布外机器人基础计划分数相对提高多达16.6%。定性双任务机器人部署研究进一步说明了基于记忆的语言计划如何与下游控制交互，总体结果表明可编辑、巩固的记忆可为机器人规划提供记忆上下文。
英文摘要
Embodied agents accumulate experience over time. We study how accumulated experience can be formed into persistent memory for future reasoning and action. We formulate Embodied Action Memory (EAM) as the capability to form and use memory over embodied experience, together with the persistent memory state produced by that process. We introduce MEMORA, a framework that instantiates EAM through a formation-consolidation-retrieval lifecycle and a multi-store world-memory architecture. MEMORA organizes experience into participant-specific Environment, Entity, Activity, and Inferred Knowledge stores: online editing revises memory as new evidence arrives, while offline consolidation abstracts repeated experience into reusable routines, habits, and preferences. We evaluate MEMORA with MEMORA-Bench, a 45-hour egocentric-video suite that measures both retrospective memory faithfulness and prospective memory-grounded planning. Across four open-weight answer models, MEMORA achieves the strongest aggregate planning performance among the evaluated memory interfaces, with its largest gains on out-of-distribution planning. On these tasks, MEMORA improves Robot-Grounded Plan score by up to 16.6 percent, suggesting that memory formed and consolidated across experience can support planning for new goals beyond directly observed episodes. A physical-robot demonstration further shows that memory formed solely from human egocentric video can ground high-level robot plans in participant-specific objects and preferences. Project website: https://github.com/yuzihaowashu/MEMORA
Comments50 pages. v1: Oral presentation at the Robotics: Science and Systems 2026 Workshop on Foundation Models for Robot Planning (FM4RoboPlan). v2: Accepted to the Association for Computational Linguistics: EMNLP 2026
