Welcome to the Autonomous Agents and Robotics Research (A2R2) website!

The Autonomous Agents and Robotics Research (A2R2) Group focuses on advancing the foundations and applications of intelligent autonomous systems. Our research lies at the intersection of reinforcement learning, robotics, and multi-agent systems, with the goal of enabling agents to learn, adapt, and cooperate in complex, dynamic environments.

We develop novel algorithms for decision-making under uncertainty, leveraging modern advances in deep reinforcement learning to train agents capable of acquiring sophisticated behaviours through interaction. Our work explores both theoretical and practical challenges, including sample efficiency, stability, and generalisation to real-world robotic systems.

A key focus of the group is robot learning in embodied settings, where perception, control, and reasoning must be tightly integrated. We investigate how agents can learn motor skills, coordination strategies, and task representations that transfer from simulation to physical platforms. This direction is motivated by the need for robots to interact naturally with humans and operate robustly in real-world settings.

A2R2 main research interests are:

• Artificial Intelligence, Artificial Neural Networks, Machine Learning, Cognitive and Developmental Robotic, Bioinspired Models, Explainable Artificial Intelligence
• Reinforcement Learning, Contextual Affordances, Dynamic Models, Grey Box Neural Models
• Interactive Reinforcement Learning, Explainable Reinforcement Learning, Human-Robot Interaction, Multimodal Integration