gym-pybullet-drones

This is a minimalist refactoring of the original gym-pybullet-drones repository, designed for compatibility with gymnasium, stable-baselines3 2.0, and SITL betaflight/crazyflie-firmware.

NOTE: if you prefer to access the original codebase, presented at IROS in 2021, please git checkout [paper|master] after cloning the repo, and refer to the corresponding README.md's.

Installation

Tested on Intel x64/Ubuntu 22.04 and Apple Silicon/macOS 14.1.

git clone https://github.com/utiasDSL/gym-pybullet-drones.git
cd gym-pybullet-drones/

conda create -n drones python=3.10
conda activate drones

pip3 install --upgrade pip
pip3 install -e . # if needed, `sudo apt install build-essential` to install `gcc` and build `pybullet`

Use

PID control examples

cd gym_pybullet_drones/examples/
python3 pid.py # position and velocity reference
python3 pid_velocity.py # desired velocity reference

Downwash effect example

cd gym_pybullet_drones/examples/
python3 downwash.py

Reinforcement learning examples (SB3's PPO)

cd gym_pybullet_drones/examples/
python learn.py # task: single drone hover at z == 1.0
python learn.py --multiagent true # task: 2-drone hover at z == 1.2 and 0.7

utiasDSL pycffirmware Python Bindings example (multiplatform, single-drone)

Install pycffirmware for Ubuntu, macOS, or Windows

cd gym_pybullet_drones/examples/
python3 cff-dsl.py

Betaflight SITL example (Ubuntu only)

git clone https://github.com/betaflight/betaflight # use the `master` branch at the time of writing (future release 4.5)
cd betaflight/ 
make arm_sdk_install # if needed, `apt install curl``
make TARGET=SITL # comment out line: https://github.com/betaflight/betaflight/blob/master/src/main/main.c#L52
cp ~/gym-pybullet-drones/gym_pybullet_drones/assets/eeprom.bin ~/betaflight/ # assuming both gym-pybullet-drones/ and betaflight/ were cloned in ~/
betaflight/obj/main/betaflight_SITL.elf

In another terminal, run the example

conda activate drones
cd gym_pybullet_drones/examples/
python3 beta.py --num_drones 1 # check the steps in the file's docstrings to use multiple drones

Citation

If you wish, please cite our IROS 2021 paper (and original codebase) as

@INPROCEEDINGS{panerati2021learning,
      title={Learning to Fly---a Gym Environment with PyBullet Physics for Reinforcement Learning of Multi-agent Quadcopter Control}, 
      author={Jacopo Panerati and Hehui Zheng and SiQi Zhou and James Xu and Amanda Prorok and Angela P. Schoellig},
      booktitle={2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
      year={2021},
      volume={},
      number={},
      pages={7512-7519},
      doi={10.1109/IROS51168.2021.9635857}
}

References

• Carlos Luis and Jeroome Le Ny (2016) Design of a Trajectory Tracking Controller for a Nanoquadcopter
• Nathan Michael, Daniel Mellinger, Quentin Lindsey, Vijay Kumar (2010) The GRASP Multiple Micro UAV Testbed
• Benoit Landry (2014) Planning and Control for Quadrotor Flight through Cluttered Environments
• Julian Forster (2015) System Identification of the Crazyflie 2.0 Nano Quadrocopter
• Antonin Raffin, Ashley Hill, Maximilian Ernestus, Adam Gleave, Anssi Kanervisto, and Noah Dormann (2019) Stable Baselines3
• Guanya Shi, Xichen Shi, Michael O’Connell, Rose Yu, Kamyar Azizzadenesheli, Animashree Anandkumar, Yisong Yue, and Soon-Jo Chung (2019) Neural Lander: Stable Drone Landing Control Using Learned Dynamics
• C. Karen Liu and Dan Negrut (2020) The Role of Physics-Based Simulators in Robotics
• Yunlong Song, Selim Naji, Elia Kaufmann, Antonio Loquercio, and Davide Scaramuzza (2020) Flightmare: A Flexible Quadrotor Simulator

Core Team WIP

• Multi-drone crazyflie-firmware SITL support (@spencerteetaert, @JacopoPan)
• Use SITL services with steppable simulation (@JacopoPan)

Desired Contributions/PRs

• Add motor delay, advanced ESC modeling by implementing a buffer in BaseAviary._dynamics()
• Replace rpy with quaternions (and ang_vel with body rates) by editing BaseAviary._updateAndStoreKinematicInformation(), BaseAviary._getDroneStateVector(), and the .computeObs() methods of relevant subclasses

Troubleshooting

• On Ubuntu, with an NVIDIA card, if you receive a "Failed to create and OpenGL context" message, launch nvidia-settings and under "PRIME Profiles" select "NVIDIA (Performance Mode)", reboot and try again.

Run all tests from the top folder with

pytest tests/

---

University of Toronto's Dynamic Systems Lab / Vector Institute / University of Cambridge's Prorok Lab