Decentralized Multi-Robot Relative Navigation in Unknown, Structurally Constrained Environments under Limited Communication (2510.09188v1)

Abstract: Multi-robot navigation in unknown, structurally constrained, and GPS-denied environments presents a fundamental trade-off between global strategic foresight and local tactical agility, particularly under limited communication. Centralized methods achieve global optimality but suffer from high communication overhead, while distributed methods are efficient but lack the broader awareness to avoid deadlocks and topological traps. To address this, we propose a fully decentralized, hierarchical relative navigation framework that achieves both strategic foresight and tactical agility without a unified coordinate system. At the strategic layer, robots build and exchange lightweight topological maps upon opportunistic encounters. This process fosters an emergent global awareness, enabling the planning of efficient, trap-avoiding routes at an abstract level. This high-level plan then inspires the tactical layer, which operates on local metric information. Here, a sampling-based escape point strategy resolves dense spatio-temporal conflicts by generating dynamically feasible trajectories in real time, concurrently satisfying tight environmental and kinodynamic constraints. Extensive simulations and real-world experiments demonstrate that our system significantly outperforms in success rate and efficiency, especially in communication-limited environments with complex topological structures.

• The paper introduces a decentralized, hierarchical framework that integrates topological mapping with real-time conflict resolution for efficient navigation.
• The paper employs lightweight topological maps derived from lidar data to achieve shared global awareness without heavy communication overhead.
• The paper validates its approach through simulations and real-world experiments, demonstrating superior success rates and dynamic safety in complex environments.

Decentralized Multi-Robot Relative Navigation in Unknown, Structurally Constrained Environments under Limited Communication

Introduction

Multi-robot navigation in unknown environments with structural constraints presents significant challenges due to limited communication, requiring a balance between global strategic planning and local tactical agility. Centralized systems achieve optimal path planning but suffer from high communication overhead, while distributed systems offer better efficiency but lack comprehensive environmental awareness. This paper introduces a decentralized, hierarchical framework that integrates strategic foresight with tactical agility without relying on a unified coordinate system. The proposed system utilizes lightweight topological maps exchanged during robot encounters, fostering emergent global awareness and facilitating efficient route planning. Each robot also engages in a sampling-based approach to resolve spatio-temporal conflicts dynamically, enhancing navigation efficiency and safety.

Hierarchical Framework Design

The proposed framework consists of two main layers: the high-level Topology Sharing and Global Guidance (TSGG) layer, and the low-level Conflict Resolution and Trajectory Planning (CRTP) layer.

Figure 1: Our hierarchical framework enables a team of robots to navigate efficiently in topologically complex environments with limited communication.

Topology Sharing and Global Guidance

In the TSGG layer, robots construct and exchange topological maps based on a visibility graph, providing a robust abstraction of the navigable space. This approach circumvents the need for dense metric data by employing sparse local exchanges, which culminates in a shared environmental understanding that enhances strategic path planning.

Figure 2: Overview of our proposed decentralized, hierarchical navigation framework.

Conflict Resolution and Trajectory Planning

The CRTP layer translates global strategic guidance into real-time collision-free trajectories. It employs a reactive approach incorporating an escape point strategy to preemptively resolve potential deadlocks. This ensures the dynamic feasibility and local safety of planned trajectories, while strategically aligning them with long-term objectives from the TSGG layer.

Figure 3: The local conflict resolution process.

Methodology

Topology Sharing Process: Lightweight topological maps are constructed by projecting lidar point clouds into binary images, extracting contours, and fusing disparate observations. This decentralized approach allows robots to maintain a comprehensive global map without overwhelming communication channels.

Conflict Resolution Process: Using the strategic route as reference, the system dynamically samples escape points upon detecting inaccessible paths, optimizing trajectory refinements through nonlinear programming to ensure collision avoidance and adherence to kinodynamic constraints.

Figure 4: An example of the system in action.

Experimental Validation

Experiments were conducted in both simulated and real-world environments to assess the system's performance across multiple scenarios featuring complex obstacles and communication limitations. The results indicated superior success rates compared to state-of-the-art methods, validating the efficiency and robustness of the proposed framework.

Figure 5: Qualitative results from three challenging simulation scenarios.

Figure 6: Real-world experiment with 3 robots in a corridor environment.

Implications and Future Work

The implications of this decentralized navigation framework are significant, offering a scalable and efficient approach for multi-robot systems in GPS-denied and structurally complex environments. Future developments will focus on scaling the system for larger robot teams and integrating richer semantic information to expand functionality beyond navigation.

Conclusion

This paper presents a decentralized hierarchical navigation system that addresses the limitations of existing methods by synergizing strategic topological guidance with tactical metric-based collision avoidance. The framework demonstrates resilience in communication-constrained environments, offering robust and efficient navigation solutions for multi-robot systems.