The lab

Learning-based control and coordination strategies for large-scale, heterogeneous multi-agent systems using compositional methods.

Multi-robot decision-making systems for high-level task instructions using distributed algorithms.

Sim-to-real transfer using MARL and GNNs, with focus on interpretability and resilience. Best Paper Award at ICRA 2023.

Diversity-seeking learning algorithms using reinforcement learning and evolutionary computation for robotic system resilience.

Bio-inspired vision-based perception and control for drone swarms navigating constrained environments.

Scalability and provable safety guarantees in multi-agent systems and swarm robotics, with interest in alignment.

Intersection of mathematics, ML theory, and robotics — parameterising uncertainty and resilience in multi-agent systems.

Wireless data communications and multi-robot control algorithms for drone swarms and automotive networks.

Moving beyond monolithic AI toward specialised, collaborative robot teams — bridging theory with real-world drone experiments.

Making GNNs safer with emphasis on robustness and interpretability in multi-agent contexts.

Multi-agent system-environment relationships, focusing on reinforcement learning and graph neural networks.

Multi-agent interactive systems using MARL and control — heterogeneity and swarm robotics implementation.

Enabling collaborative robot team reasoning — zero-shot coordination and world models.

MARL representing agent intentions without communication — exploring heterogeneity in coordination.

Coordinates lab space, events, resources, and grant administration.

Moving cutting-edge research to real-world applications across computer vision, imitation learning, and RL.