GUIDE: A Diffusion-Based Autonomous Robot Exploration Framework Using Global Graph Inference

2025-09-24

AI Summary

To address the challenge of modeling unobserved regions and planning globally efficient paths for autonomous exploration in complex indoor environments, this paper proposes a novel exploration framework integrating global graph reasoning with diffusion-based decision-making. Methodologically, it employs graph neural networks for environment modeling and introduces a region-level evaluation mechanism to enhance prediction fidelity. Its key contributions are: (1) a region-assessment-oriented global graph representation that jointly leverages structured inference and uncertainty-aware modeling to improve reliability in unknown-space characterization; and (2) a low-step diffusion policy network that generates anticipatory action sequences, enabling stable and efficient end-to-end decision-making. Extensive experiments in both simulation and real-world settings demonstrate that the method achieves an 18.3% faster coverage completion rate and reduces redundant motion by 34.9% compared to state-of-the-art approaches, significantly improving exploration efficiency and robustness.

Technology Category

• Planning, Routing, and Scheduling: Model-Based Reasoning
• Search and Optimization: Learning to Search
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Content-based information diffusion
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Autonomous exploration in structured and complex indoor environments remains a challenging task, as existing methods often struggle to appropriately model unobserved space and plan globally efficient paths. To address these limitations, we propose GUIDE, a novel exploration framework that synergistically combines global graph inference with diffusion-based decision-making. We introduce a region-evaluation global graph representation that integrates both observed environmental data and predictions of unexplored areas, enhanced by a region-level evaluation mechanism to prioritize reliable structural inferences while discounting uncertain predictions. Building upon this enriched representation, a diffusion policy network generates stable, foresighted action sequences with significantly reduced denoising steps. Extensive simulations and real-world deployments demonstrate that GUIDE consistently outperforms state-of-the-art methods, achieving up to 18.3% faster coverage completion and a 34.9% reduction in redundant movements.

Problem

• Improving autonomous robot exploration in complex indoor environments
• Addressing limitations in modeling unobserved space and global path planning
• Reducing inefficient movements and accelerating area coverage completion

Innovation

• Global graph inference with region-evaluation mechanism
• Diffusion policy network for stable action sequences
• Integrates observed data and unexplored area predictions