Learning Decentralized Swarms Using Rotation Equivariant Graph Neural Networks

2025-02-24

AI Summary

This work addresses the decentralized cooperative control of swarm agents lacking a central coordinator. Method: We propose a graph neural network (GNN)-based controller that jointly enforces rotational equivariance and translational invariance—geometric symmetries explicitly embedded into a decentralized GNN architecture for the first time. Our approach integrates symmetry-constrained modeling with decentralized multi-agent reinforcement learning to achieve distributed control that simultaneously ensures formation stability and robustness. Contribution/Results: Compared to baseline GNN controllers, our method reduces training data requirements by 70% and trainable parameters by 75%, while significantly improving generalization across unseen communication topologies, swarm scales, and external disturbances. The resulting symmetry-aware control paradigm enables efficient, scalable coordination for applications such as autonomous maritime fleet navigation and dynamic monitoring in large-scale sensor networks.

Technology Category

• Multiagent Systems: Coordination and Collaboration
• Intelligent Robots: Behavior Learning & Control
• Machine Learning: Deep Neural Architectures and Foundation Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Decentralized Web and Fediverse systems
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

The orchestration of agents to optimize a collective objective without centralized control is challenging yet crucial for applications such as controlling autonomous fleets, and surveillance and reconnaissance using sensor networks. Decentralized controller design has been inspired by self-organization found in nature, with a prominent source of inspiration being flocking; however, decentralized controllers struggle to maintain flock cohesion. The graph neural network (GNN) architecture has emerged as an indispensable machine learning tool for developing decentralized controllers capable of maintaining flock cohesion, but they fail to exploit the symmetries present in flocking dynamics, hindering their generalizability. We enforce rotation equivariance and translation invariance symmetries in decentralized flocking GNN controllers and achieve comparable flocking control with 70% less training data and 75% fewer trainable weights than existing GNN controllers without these symmetries enforced. We also show that our symmetry-aware controller generalizes better than existing GNN controllers. Code and animations are available at http://github.com/Utah-Math-Data-Science/Equivariant-Decentralized-Controllers.

Problem

• Decentralized swarm control optimization
• Maintaining flock cohesion challenges
• Symmetry-aware GNN controller generalization

Innovation

• Rotation Equivariant Graph Networks
• Decentralized Flocking Control
• Reduced Training Data Usage