HIVEX

Multi-Agent Reinforcement Learning environment suite with a focus on critical ecological challenges.

About

The motivation of the hivex suite is to provide advanced reinforcement learning benchmarking environments with an emphasis on: (1) real-world scenarios, (2) multi-agent systems, (3) investigating problems and solutions with high impact on society.

Train & Test Scripts: https://github.com/hivex-research/hivex

Available Environments: https://github.com/hivex-research/hivex-environments

Wind Farm Control	Wildfire Resource Management	Drone-Based Reforestation	Ocean Plastic Collection	Aerial WildFire Suppression
WFC	WRM	DBR	OPC	AWS
Task Count / Patterns	Task Count / Difficulty Levels	Task Count / Difficulty Levels	Task Count / Difficulty Levels	Task Count / Difficulty Levels
2 / 9	3 / 10	7 / 10	4 / -	9 / 10