Graph-Based Topological Exploration Planning in Large-Scale 3D Environments

Abstract

Currently, state-of-the-art exploration methods make various efforts in constructing and maintaining high-resolution world representations to acquire positions in configuration space that maximize information gain. However, those “optimal” selections could quickly become obsolete due to the influx of new information, especially in large-scale environments, which results in high-frequency re-planning that hinders the exploration efficiency. In this paper, we propose a graph-based topological planning framework, building a sparse topological map in three-dimensional (3D) space to guide exploration steps with high-level intents so as to render consistent exploration maneuvers. Specifically, this work presents a novel method to represent 3D spaces as convex polyhedrons, whose geometry information is utilized to group spaces into distinctive regions. These distinctive regions are then added as nodes into the topological map, directing the exploration process. We compared our method with the state-of-art in simulated environments. The proposed method achieves better exploration performance in space coverage and outperforms exploration efficiency by more than 40%. Finally, a field experiment was conducted to further evaluate the applicability of our method to empower efficient and robust exploration in real-world environments.