Davide Domini

Azienda-Unita' Sanitaria Locale Di Cesena, University of Bologna

Papers

Total Citations

H-Index

About

Davide Domini is a rising researcher in multi-agent reinforcement learning (MARL) and swarm robotics, focusing on the critical challenge of scaling coordination policies for large groups of autonomous agents. His work addresses the fundamental tension between decentralized decision-making and collective intelligence, particularly in domains like robotics, traffic management, and IoT systems. Domini’s major contributions include developing neighbor-based training strategies that enable agents to learn from local interactions without centralized oversight, and pioneering the use of graph neural networks (GNNs) to scale swarm coordination—exploring how far these architectures can push performance as agent numbers grow. His 2025 paper on neighbor-based decentralized training (3 citations) and his GNN scalability study (2 citations) represent early but impactful steps toward practical many-agent systems. Domini also created a reusable simulation pipeline for many-agent reinforcement learning (2024, 2 citations), providing a standardized platform for testing and comparing coordination policies. With a demonstrator for self-organizing robot teams (2025), he bridges theory and application, showcasing how his algorithms enable real-world robotic swarms to achieve emergent, scalable behaviors. His work is shaping the future of autonomous systems that must operate reliably in large, dynamic environments.