RoboAtlas: Contextual Active SLAM

2026-06-24

AI Summary

This work addresses the challenge of adaptively balancing geometric exploration and semantic reasoning in active SLAM within large-scale, complex environments. The authors propose RoboAtlas, a novel framework that integrates the scalable 3D semantic mapping system OpenRoboVox with context-aware active SLAM. RoboAtlas employs a contextual multi-armed bandit mechanism to dynamically fuse frontier-based exploration, global semantic map reasoning, and egocentric visual-language model inference (e.g., GPT-4o and Qwen2.5-VL-7B), enabling semantic-driven exploration–exploitation trade-offs. Experimental results demonstrate that the method achieves 100% task success in a real-world 1800 m² environment. On the GOAT-Bench “Val Unseen” benchmark, it attains a 90.6% success rate with GPT-4o—outperforming the strongest baseline by 17.8 percentage points—and 88.8% with Qwen2.5-VL-7B, surpassing all GPT-4o-based baselines.

Abstract

We present RoboAtlas, a contextual Active SLAM framework that adaptively balances geometric exploration and semantic reasoning using a scalable 3D semantic mapping system, OpenRoboVox. RoboAtlas integrates frontier exploration, global semantic-map reasoning, and egocentric VLM-based reasoning through a contextual multi-armed bandit that transitions from exploration to semantically guided navigation as scene understanding improves. We evaluate the system in simulation and on a Unitree Go2 robot in large-scale real-world environments exceeding 1800 m2 with approx. 30k mapped semantic instances, achieving a 100% task success rate. On the GOAT-Bench "Val Unseen" benchmark, RoboAtlas achieves state-of-the-art performance with highest reported success rate (SR) of 90.6%, using GPT-4o, improving over the strongest prior baseline by 17.8 percentage points in SR. Using the much smaller Qwen2.5-VL-7B model, it still achieves 88.8% SR, outperforming all baselines using GPT-4o in SR, and revealing the importance of the information gained by our semantic mapping framework over simply replacing the underlying foundation model. The results demonstrate that grounding foundation models with large-scale 3D semantic maps enables robust and efficient contextual Active SLAM.

Problem

• Active SLAM
• semantic mapping
• contextual reasoning
• exploration-exploitation trade-off
• 3D scene understanding

Innovation

• Contextual Active SLAM
• 3D Semantic Mapping
• Multi-armed Bandit
• Vision-Language Model (VLM)
• Semantic-guided Navigation