Scalable Multi-Agent Maze Traversal with Local Communication

2026-08-12

AI Summary

This work addresses the challenge of efficiently coordinating multiple agents with only local communication to achieve collision-free target localization and complete traversal in an unknown maze-like environment. The authors propose a distributed algorithm that employs a dynamically switching leader-follower mechanism, wherein a single agent performs exploration while the others collaboratively avoid collisions and propagate local information, thereby enabling collective traversal of the unknown graph structure. This approach constitutes the first provably complete, scalable multi-agent traversal strategy relying solely on local communication, with makespan asymptotically matching that of the omniscient optimal solution as the number of agents increases. Theoretical analysis establishes the algorithm’s completeness and bounded complexity, while simulations demonstrate its scalability to up to 625 agents and show that average total energy consumption decreases with system size, significantly outperforming baseline methods where agents operate independently.

Abstract

Cave networks, pipe systems, and similar maze-like environments pose significant challenges for multi-agent navigation in unknown settings with limited communication. We propose a distributed algorithm that enables agents to collectively traverse an unknown, possibly cyclic graph. Agents enter sequentially at a designated start node and are tasked to localize and reach an undisclosed goal while avoiding collisions. They coordinate via local communication using leader-follower relationships and leader switching. At any moment in time, exploration is performed by only one of the agents, which runs a single-agent maze solver. We prove that the algorithm is complete, that its makespan is asymptotically equivalent (in the number of agents) to that of an optimal full-knowledge strategy, and derive its time and space complexity. Simulations with up to $625$ agents show a decreasing average sum-of-fuels as the number of agents increases and demonstrate that the proposed approach outperforms a naïve baseline in which all agents independently execute the single-agent solver.

Problem

• multi-agent navigation
• maze traversal
• local communication
• unknown environments
• collision avoidance

Innovation

• multi-agent coordination
• local communication
• leader-follower dynamics
• distributed maze traversal
• asymptotically optimal makespan