Supaero Reinforcement Learning Initiative

View our GitHub profile

The SuReLI blog

The SuReLIb (private)

Software projects

The software section of this site links to open source code developped during projects stemming from our research interests. We generally open source our code when we believe it has a value to the community. Consequently, if some research area seems to lack open source code, there probably is something cooking internally.

Deep RL agents

Our library of Deep RL agents. A great place to get started for most people who work with SuReLI (inside or outside collaborators).

Contributors: Valentin Guillet, Emmanuel Rachelson.

Dynamic Gym

This is a pip package implementing Reinforcement Learning algorithms in non-stationary environments supported by the OpenAI Gym toolkit.

Contributor: Erwan Lecarpentier.

Multitask Acrobot

A Gym-compatible acrobot environment with several reward functions corresponding to different tasks.

Contributor: Jean-Jacques Simeoni.


Naive Bayes Classification for Subset Selection (NaiBX): an extension of Naive Bayes to multi-label classification.

See also the related paper: Naive Bayes Classification for Subset Selection

Contributors: Luca Mossina, Emmanuel Rachelson.

iBoat stall avoidance

Using Deep RL to avoid or recover from stall on an autonomous sail-boat’s wing.

Contributors: N. Megel, A. Bonet-Munoz, T. Karch, Y. Brière, E. Rachelson.

iBoat navigation planning

MCTS planning for the long-term path planning of an autonomous sail-boat.

Contributors: F. Brulport, J.-M. Belley, P. Barde, C. Chanel, Y. Brière, E. Rachelson.

Learning to fly - the simulator

A stand-alone, C++ simulator of the flight dynamics of an autonomous glider within convective soaring conditions. Used to benchmark RL methods for control and planning of autonomous gliders.

Contributors: S. Rapp, M. Melo Oliver, R. Madelaine, L. Becq, A. Bufort, H. Akhmouch, T. Le Minh, E. Herlaut, N. El Jaafari, S. Ganapathi-Raju, E. Lecarpentier, E. Rachelson.