![[2510.09469] Towards Information-Optimized Multi-Agent Path Finding: A Hybrid Framework with Reduced Inter-Agent Information Sharing](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2510.09469] Towards Information-Optimized Multi-Agent Path Finding: A Hybrid Framework with Reduced Inter-Agent Information Sharing
Summary
This paper presents a hybrid framework for Multi-Agent Path Finding (MAPF) that minimizes inter-agent information sharing while maintaining solution quality, addressing scalability and privacy concerns in robotics.
Why It Matters
The research addresses a critical challenge in robotics and autonomous systems by proposing a solution that balances the need for effective communication among agents with the constraints of privacy and bandwidth. This is particularly relevant as the use of autonomous systems expands in various applications, necessitating efficient and secure communication protocols.
Key Takeaways
- Introduces a hybrid framework, IO-MAPF, for efficient MAPF solutions.
- Achieves significant reductions in information sharing (2x to 23x) without compromising success rates.
- Utilizes reinforcement learning for decentralized planning, supported by a centralized coordinator.
- Addresses privacy and bandwidth issues in multi-agent systems.
- Demonstrates effectiveness through simulation and hardware experiments.
Computer Science > Multiagent Systems arXiv:2510.09469 (cs) [Submitted on 10 Oct 2025 (v1), last revised 22 Feb 2026 (this version, v2)] Title:Towards Information-Optimized Multi-Agent Path Finding: A Hybrid Framework with Reduced Inter-Agent Information Sharing Authors:Bharath Muppasani, Ritirupa Dey, Biplav Srivastava, Vignesh Narayanan View a PDF of the paper titled Towards Information-Optimized Multi-Agent Path Finding: A Hybrid Framework with Reduced Inter-Agent Information Sharing, by Bharath Muppasani and 3 other authors View PDF HTML (experimental) Abstract:Multi-agent pathfinding (MAPF) remains a critical problem in robotics and autonomous systems, where agents must navigate shared spaces efficiently while avoiding conflicts. Traditional centralized algorithms with global information provide high-quality solutions but scale poorly in large-scale scenarios due to the combinatorial explosion of conflicts. Conversely, distributed approaches that have local information, particularly learning-based methods, offer better scalability by operating with relaxed information availability, yet often at the cost of solution quality. In realistic deployments, information is a constrained resource: broadcasting full agent states and goals can raise privacy concerns, strain limited bandwidth, and require extra sensing and communication hardware, increasing cost and energy use. We focus on the core question of how MAPF can be solved with minimal inter-agent information sharing whi...