ABot-AgentOS: A General Robotic Agent OS with Lifelong Multi-modal Memory

2026-07-11

AI Summary

This work addresses the lack of a unified runtime system in existing long-horizon embodied agents, which hinders coordinated reasoning, memory management, tool invocation, and cross-platform execution. The authors propose ABot-AgentOS—a general-purpose robotic agent operating system situated above low-level controllers—that enables context-aware planning, context-isolated skill execution, multi-stage verification, multimodal memory, and edge-cloud collaboration. Key innovations include a universal agent OS architecture, a Universal Multi-modal Graph Memory, and a failure-driven self-evolution mechanism that ensures data isolation while enabling continuous performance improvement. Experiments demonstrate that the system outperforms single-controller baselines on a subset of EmbodiedWorldBench and achieves state-of-the-art results on memory-centric benchmarks including LoCoMo, OpenEQA, Mem-Gallery, and NExT-QA, with the self-evolution mechanism further enhancing multiple evaluation metrics.

Abstract

Recent VLM and VLA systems have improved robotic perception and action prediction, yet long-horizon embodied agents still require a general runtime layer for reasoning, memory, tool use, verification, and cross-embodiment execution. We present ABot-AgentOS, a general robotic Agent Operating System that sits above low-level controllers and provides a deliberative agent layer for scene-conditioned planning, context-isolated skill execution, multi-stage verification, multi-modal memory, and edge-cloud collaboration. To evaluate such systems, we introduce EmbodiedWorldBench, an executable benchmark with 16 indoor, outdoor, and hybrid scenes, four difficulty levels, and over 200 tasks involving navigation, object search, NPC dialogue, dynamic events, and trace-grounded scoring. ABot-AgentOS further introduces Universal Multi-modal Graph Memory, a persistent source-grounded substrate that converts dialogue, visual observations, spatial context, temporal relations, and task traces into typed nodes and edges. A failure-driven self-evolution loop converts diagnosed memory failures into gated runtime evo-assets that are promoted only to later evaluation splits, preventing current-split ground-truth leakage while enabling continual improvement. On an initial EmbodiedWorldBench subset, ABot-AgentOS improves over a single-controller baseline in both task success and goal completion. Across memory benchmarks, ABot-AgentOS Static achieves 87.5 on LoCoMo, 59.9 on OpenEQA EM-EQA, 88.6 on Mem-Gallery, and 76.5 Acc@All on NExT-QA; self-evolution further improves LoCoMo to 88.7, OpenEQA to 60.4, and Mem-Gallery to 89.0. These results suggest that a general Agent OS layer can improve long-horizon embodied execution while providing persistent, auditable memory for continual interaction.

Problem

• embodied agents
• long-horizon tasks
• multi-modal memory
• robotic agent OS
• cross-embodiment execution

Innovation

• Agent Operating System
• Multi-modal Memory
• Failure-driven Self-evolution
• Embodied AI
• Graph-based Memory