Perception Action Communication

Perception-action-communication (PAC) research focuses on how agents perceive their environment, communicate with others, and act accordingly, aiming to achieve coordinated behavior in complex systems. Current research emphasizes decentralized approaches, particularly leveraging graph neural networks (GNNs) to manage information flow and action selection in multi-agent systems, often within asynchronous frameworks to better reflect real-world scenarios. These models are being applied to diverse problems, including robot swarm control and synchronization of coupled oscillators, demonstrating the broad applicability of PAC principles to both theoretical and practical challenges in distributed systems.