RoboMemory: A Brain-inspired Multi-memory Agentic Framework for Interactive Environmental Learning in Physical Embodied Systems
cs.RO, cs.AI
Submitted: 2025-08-02
Updated: 2026-11-10
Project page: https://sp4595.github.io/robomemory
License: http://creativecommons.org/licenses/by-nc-sa/4.0/
The gist: Long-horizon embodied tasks require agents to learn, reason, and generalize robustly across complex real-world environments.
Terminology
Abstract
Long-horizon embodied tasks require agents to learn, reason, and generalize robustly across complex real-world environments. However, existing approaches often struggle with partial observability, fragmented spatial reasoning, and inefficient integration of heterogeneous memories, limiting their capacity for long-horizon adaptation. To address this, we introduce RoboMemory, a brain-inspired framework that unifies Temporal, Spatial, Semantic, and Episodic memory for efficient long-horizon planning and interactive learning. Its core innovations are a comprehensive memory system designed for embodied tasks, and a dynamic spatial knowledge graph for scalable, consistent memory updates. Extensive experiments on EmbodiedBench show that RoboMemory improves the average success rate of Qwen2.5-VL-72B-Ins by 26.5 percentage points and achieves performance comparable to strong closed-source VLM agents. Real-world trials further suggest RoboMemory's potential for interactive environmental learning by reusing accumulated experience across repeated task attempts. These results demonstrate RoboMemory as a scalable memory-augmented framework for embodied agents, improving long-horizon planning and cumulative learning in complex interactive environments.
Sources
- Agent S: An Open Agentic Framework that Uses Computers Like a Human
- Qwen2.5-VL Technical Report
- GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
- $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
- GOAT: GO to Any Thing
- Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
- Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory
- NeSyC: A Neuro-symbolic Continual Learner For Complex Embodied Tasks In Open Domains
- MSI-Agent: Incorporating Multi-Scale Insight into Embodied Agents for Superior Planning and Decision-Making
- Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body Teleoperation
- LLM-Empowered Embodied Agent for Memory-Augmented Task Planning in Household Robotics
- HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models
- Toward General-Purpose Robots via Foundation Models: A Survey and Meta-Analysis
- GPT-4o System Card
- OpenVLA: An Open-Source Vision-Language-Action Model
- CLEA: Closed-Loop Embodied Agent for Enhancing Task Execution in Dynamic Environments
- Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems
- MemGPT: Towards LLMs as Operating Systems
- Hi Robot: Open-Ended Instruction Following with Hierarchical Vision-Language-Action Models
- RoboOS: A Hierarchical Embodied Framework for Cross-Embodiment and Multi-Agent Collaboration
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation