A Hierarchical Region-Based Approach for Efficient Multi-Robot Exploration

TL;DR

This work tackles efficient unknown-environment exploration by introducing RegionGraph, a spatially informative representation of unexplored regions, and a two-tier hierarchical framework that enables asynchronous global planning and local refinement. The RegionGraph supports a VRP-based task allocation that preserves spatial structure, while the center/client architecture reduces global planning frequency and enhances robustness. Empirical results in simulations and real-world robot trials show substantial improvements in exploration time, total distance traveled, and overlap reduction, illustrating practical impact for multi-robot systems. The approach offers a scalable path toward coordinated, efficient exploration, with potential extensions to 3D environments.

Abstract

Multi-robot autonomous exploration in an unknown environment is an important application in robotics.Traditional exploration methods only use information around frontier points or viewpoints, ignoring spatial information of unknown areas. Moreover, finding the exact optimal solution for multi-robot task allocation is NP-hard, resulting in significant computational time consumption. To address these issues, we present a hierarchical multi-robot exploration framework using a new modeling method called RegionGraph. The proposed approach makes two main contributions: 1) A new modeling method for unexplored areas that preserves their spatial information across the entire space in a weighted graph called RegionGraph. 2) A hierarchical multi-robot exploration framework that decomposes the global exploration task into smaller subtasks, reducing the frequency of global planning and enabling asynchronous exploration. The proposed method is validated through both simulation and real-world experiments, demonstrating a 20% improvement in efficiency compared to existing methods.

Figures (12)

• Figure 1: Results of a task allocation (left) and the explored map (right) in the real-world experiment. The image on the left displays a planning result on a global scale. The colored areas represent partitioned regions. The gray area represents uninterested areas outside the task.
• Figure 2: System Overview
• Figure 3: The Exploration Algorithm Using the RegionGraph. (a) Map updates received. (b) Unexplored areas are first partitioned using a Voronoi-like method. (c) Then, a RegionGraph is constructed, with each vertex representing a region and each edge representing a connection. (d) Finally, the optimal visiting sequences can be obtained after solving the VRP.
• Figure 4: The Hierarchical Framework. The center node assigns subtasks to the client nodes, as shown in \ref{['hierarchical1']}. In \ref{['hierarchical2']}, a client node receives the task and transfer along the guide path to the first region. \ref{['hierarchical3']}, \ref{['hierarchical4']}, and \ref{['hierarchical5']} are the map within the red dashed box in \ref{['hierarchical2']}. The client node partitions the unexplored areas covered by ROI into regions, as shown in \ref{['hierarchical3']}. It then constructs a RegionGraph in \ref{['hierarchical4']} and solves the TSP to generate the optimal visiting sequence before navigating to the first target, as shown in \ref{['hierarchical5']}.
• Figure 5: The State Diagram of the Hierarchical Framework