OpenaAI Gym Franka Emika Panda robot environment based on PyBullet.
pip install panda-gymgit clone https://github.com/qgallouedec/panda-gym.git
pip install -e panda-gymimport gym
import panda_gym
env = gym.make('PandaReach-v1', render=True)
obs = env.reset()
done = False
while not done:
action = env.action_space.sample() # random action
obs, reward, done, info = env.step(action)
env.close()You can also
PandaReach-v1 |
PandaPush-v1 |
 |
 |
PandaSlide-v1 |
PandaPickAndPlace-v1 |
 |
 |
PandaStack-v1 |
|
 |
Baselines results and pre-trained agents available in rl-baselines3-zoo.
Cite as
@misc{gallouédec2021multigoal,
title={Multi-Goal Reinforcement Learning environments for simulated Franka Emika Panda robot},
author={Quentin Gallou{\'e}dec and Nicolas Cazin and Emmanuel Dellandr{\'e}a and Liming Chen},
year={2021},
eprint={2106.13687},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
Environments are widely inspired from OpenAI Fetch environments.