Multi-Robot Rendezvous in Unknown Environment with Limited Communication

TL;DR

The paper tackles rendezvous for multi-robot systems in unknown environments under limited communication by introducing PIER, a two-step framework that combines partitioned NBV-based exploration with RP estimation and a subsequent rendezvous-point selection on merged topological maps. It leverages lightweight FHT-Maps to enable RP estimation with minimal data exchange and uses Voronoi-based space partitioning and PGO for map alignment, culminating in a sampling-based, sub-optimal rendezvous point selection. The primary contributions are the PIER algorithm, the lightweight FHT-Map for RP estimation and rendezvous point selection, and a comprehensive evaluation showing faster exploration, reduced data transmission, and near-optimal rendezvous points compared to baselines. The approach offers practical impact for scalable, bandwidth-constrained multi-robot deployments in unknown environments.

Abstract

Rendezvous aims at gathering all robots at a specific location, which is an important collaborative behavior for multi-robot systems. However, in an unknown environment, it is challenging to achieve rendezvous. Previous researches mainly focus on special scenarios where communication is not allowed and each robot executes a random searching strategy, which is highly time-consuming, especially in large-scale environments. In this work, we focus on rendezvous in unknown environments where communication is available. We divide this task into two steps: rendezvous based environment exploration with relative pose (RP) estimation and rendezvous point selection. A new strategy called partitioned and incomplete exploration for rendezvous (PIER) is proposed to efficiently explore the unknown environment, where lightweight topological maps are constructed and shared among robots for RP estimation with very few communications. Then, a rendezvous point selection algorithm based on the merged topological map is proposed for efficient rendezvous for multi-robot systems. The effectiveness of the proposed methods is validated in both simulations and real-world experiments.

Figures (6)

• Figure 1: The framework for robots rendezvous used in this work. (a) Three robots are located in different locations. (b) PIER is performed firstly to explore the environment. During this process, each robot maintains a FHT-Map and calculates RPs. (c) When $\mathcal{G}_r$ is connected, rendezvous point is selected based on the merged FHT-Map and rendezvous is achieved.
• Figure 2: Illustration for main nodes selection. The robot's motion trajectory passes a crossroad which has higher probability of being visited by other robots.
• Figure 3: The rendezvous of three robots in office Environment. (a) Three robots are located in an unknown environment. (b) Each robot perform PIER to explore the environment, while constructing a FHT-Map. (c) $\mathcal{R}^1$ passes through a main node of $\mathcal{R}^2$, thereby merging two topological maps $\mathcal{G}_{f}^1$ and $\mathcal{G}_{f}^2$. (d) Space is partitioned between $\mathcal{R}^1$ and $\mathcal{R}^2$. Then $\mathcal{R}^2$ finds a main node of the remaining $\mathcal{R}^3$. (e) Rendezvous point $\hat{\textbf{p}}^*$ is selected using the merged FHT-Map. Then each robots performs rendezvous. (f) Rendezvous among three robots is realized in the simulation.
• Figure 4: Illustration of the selected rendezvous points. The color in the map represents $F(\textbf{p})$ at location $\textbf{p}$ obtained by A* algorithm. The red circles represent robots' positions. The yellow star indicates the global optimal points $\textbf{p}^*$. The plus sign represents the position of $\hat{\textbf{p}}^*$ based on FHT-Maps. (a) Two robots. (b) Four robots. (c) Six robots. (d) Eight robots.
• Figure 5: Comparison on time consumption and error based on different map. A near optimal solution $\hat{\textbf{p}}^*$ can be obtained on FHT-Map. Besides, if there are denser nodes in FHT-Map, a more accurate result can be obtained.