Instinct-driven Reinforcement Learning

Evolving Physical Instinct for Morphology and Control Co-Adaption (IROS 2023)

Paper: [https://ieeexplore.ieee.org/document/10342243]

Installation

1. Create a new python virtual env with python 3.6, 3.7 or 3.8 (3.8 recommended)

2. Install pytorch 1.10 with cuda-11.3:

   • pip3 install torch==1.10.0+cu113 torchvision==0.11.1+cu113 torchaudio==0.10.0+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html

3. Install Isaac Gym

   • Download and install Isaac Gym Preview 3 from https://developer.nvidia.com/isaac-gym
   • cd isaacgym/python && pip install -e .
   • Try running an example cd examples && python 1080_balls_of_solitude.py
   • For troubleshooting check docs isaacgym/docs/index.html)

4. Install RL-Games

   • https://github.com/Denys88/rl_games

Usage

Run a simple training process by the following command (set headless=True to disable the viewer):

python train.py task=Dog morph=PSOSearch train=MorphTensor headless=False

Settings

We use Hydra to manage configurations

• https://hydra.cc/docs/intro/

Following main options are currently supported:

task: Kangaroo | Raptor | Dog | 
terrain: Flat | Uphill | Downhill | 
train: SinglePPO | MorphTensor 
morph: Fix | PSOSearch

See more arguments in inrl/cfg.

Paper and Citation

If you find our work useful, please consider citing us!

@inproceedings{chen2023evolving,
  title={Evolving Physical Instinct for Morphology and Control Co-Adaption},
  author={Chen, Xinglin and Huang, Da and Li, Minglong and Cai, Yishuai and Wen, Zhuoer and Cai, Zhongxuan and Yang, Wenjing},
  booktitle={2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages={6616--6623},
  year={2023},
  organization={IEEE}
}