Miguel Duarte presented on the design of communication and control systems for swarms of aquatic surface drones. The goals of the projects are to develop a scalable and fault-tolerant network architecture for drone swarms, explore control synthesis approaches for real-world maritime missions, and conduct proof-of-concept experiments with up to 25 robots. A hybrid control approach uses evolved neural networks as behavior primitives and finite state machines to arbitrate behaviors for tasks like patrolling. Preliminary results show the approach can control large swarms of aquatic drones.