RoboMemory: A Brain-inspired Multi-memory Agentic Framework for Lifelong Learning in Physical Embodied Systems

2025-08-02

AI Summary

To address key challenges in lifelong learning for physically embodied systems—including continual adaptation, multi-memory latency, task relationship modeling, and闭环 planning deadlocks—this paper proposes a brain-inspired, four-module cognitive agent architecture. The framework parallelly integrates spatial, temporal, episodic, and semantic memory subsystems, tightly coupled with a dynamic knowledge graph to ensure cross-memory consistency and scalability. A synergistic mechanism coordinates information preprocessing, embodied lifelong memory maintenance, closed-loop planning, and low-level execution—significantly reducing inference latency while enabling long-horizon planning and online learning. Evaluated on EmbodiedBench, the approach achieves a 25% average success rate improvement over open-source baselines and surpasses closed-source SOTA models by 5%. Empirical deployment demonstrates markedly enhanced success rates on repeated tasks, validating both scalability and real-world effectiveness.

Technology Category

• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Intelligent Robots: Embodied AI
• Multiagent Systems: Multiagent Learning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We present RoboMemory, a brain-inspired multi-memory framework for lifelong learning in physical embodied systems, addressing critical challenges in real-world environments: continuous learning, multi-module memory latency, task correlation capture, and infinite-loop mitigation in closed-loop planning. Grounded in cognitive neuroscience, it integrates four core modules: the Information Preprocessor (thalamus-like), the Lifelong Embodied Memory System (hippocampus-like), the Closed-Loop Planning Module (prefrontal lobe-like), and the Low-Level Executer (cerebellum-like) to enable long-term planning and cumulative learning. The Lifelong Embodied Memory System, central to the framework, alleviates inference speed issues in complex memory frameworks via parallelized updates/retrieval across Spatial, Temporal, Episodic, and Semantic submodules. It incorporates a dynamic Knowledge Graph (KG) and consistent architectural design to enhance memory consistency and scalability. Evaluations on EmbodiedBench show RoboMemory outperforms the open-source baseline (Qwen2.5-VL-72B-Ins) by 25% in average success rate and surpasses the closed-source State-of-the-Art (SOTA) (Claude3.5-Sonnet) by 5%, establishing new SOTA. Ablation studies validate key components (critic, spatial memory, long-term memory), while real-world deployment confirms its lifelong learning capability with significantly improved success rates across repeated tasks. RoboMemory alleviates high latency challenges with scalability, serving as a foundational reference for integrating multi-modal memory systems in physical robots.

Problem

• Addresses continuous learning in physical embodied systems
• Reduces multi-module memory latency via parallel processing
• Enhances task correlation capture and loop mitigation

Innovation

• Brain-inspired multi-memory framework for lifelong learning
• Parallelized memory updates across four submodules
• Dynamic Knowledge Graph enhances memory consistency