Decentralized Multi-Agent Deep Reinforcement Learning in Swarms of Drones for Flood Monitoring
David Baldazo, Juan Parras, Santiago Zazo · 2019
Multi-Agent Deep Reinforcement Learning is becoming a promising approach to the problem of coordination of swarms of drones in dynamic systems. In particular, the use of autonomous aircraft for flood monitoring is now regarded as an economically viable option and it can benefit from this kind of automation: swarms of unmanned aerial vehicles could autonomously generate nearly real-time inundation maps that could improve relief work planning. In this work, we study the use of Deep Q-Networks (DQN) as the optimization strategy for the trajectory planning that is required for monitoring floods, we train agents over simulated floods in procedurally generated terrain and demonstrate good performance with two different reward schemes.