During the 2025 edition of the ARTD Symposium in Prague we will host a BlueSky-Gym workshop.
Are you active in the field of Air Traffic Control / Management research, and have always wondered how to apply Reinforcement Learning to some of the problems, or do you just want to learn a new method? Then join at the ARTD symposium for a BlueSky-Gym workshop.
During the workshop we will explain the basics of Reinforcement Learning and Stable-Baselines3 and guide you through the installation of both BlueSky-Gym and SB3. Then we will put theory to practice and use one of the algorithms hosted through SB3 to train an air traffic control agent on a problem of your choice.
While the agents are learning, we will go over BlueSky-Gym in more detail and learn how you can subclass the existing environments and modify them for your own needs. For example by changing the observation vector or reward function. After this final interactive part of the workshop we will have an open discussion about the future developments that can make BlueSky-Gym more useful for your own research, the future of automation in ATC and how we can ensure safety and certification.
Hopefully, by the end of the workshop, you will have learned how to train your own agent and adapt BSG for your own needs. Additionally we should have some nice ideas and directions for future development. Overall, the goal of this workshop is to get you comfortable with using RL as a potential research tool in the future!