SATA

Official Implementation for SATA: Safe and Adaptive Torque-Based Locomotion Policies Inspired by Animal Learning

Accepted at Robotics: Science and Systems (RSS) 2025

LI Peizhuo*. LI Hongyi*, Ge SUN, Jin CHENG, Xinrong YANG, Guillaume BELLEGARDA, Milad SHAFIEE, Yuhong CAO, Auke IJSPEERT, Guillaume SARTORETTI

---

🧭 Overview

SATA is a torque-based reinforcement learning framework inspired by how animals progressively acquire locomotion capabilities.
It introduces a biologically motivated growth curriculum that schedules torque limits and control frequency to evolve safely during training.

---

Training in Simulation

Pipeline to Install and Train ABS

Note: Before running our code, it's highly recommended to first play with RSL's Legged Gym version to get a basic understanding of the Isaac-LeggedGym-RslRL framework.

1. Create environment and install torch

   conda create -n xxx python=3.8  # or use virtual environment/docker
   
   pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu116  
   # used version during this work: torch==2.0.1 torchvision==0.15.2 torchaudio==2.0.2
   # for older cuda ver:
   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
   

2. Install Isaac Gym preview 4 release https://developer.nvidia.com/isaac-gym

   unzip files to a folder, then install with pip:

   cd isaacgym/python && pip install -e .

   check it is correctly installed by playing:

   cd examples && python 1080_balls_of_solitude.py

3. Clone this codebase and install our rsl_rl

   pip install -e rsl_rl

4. Install our legged_gym

   pip install -e legged_gym

   Ensure you have installed the following packages:

   • pip install numpy==1.21 (must < 1.24, >1.20)
   • pip install tensorboard
   • pip install setuptools==59.5.0
   • pip install wandb

5. Try training.

   can use "--headless" to disable gui, press "v" to pause/resume gui play.

   for go2, in SATA/legged_gym/legged_gym/envs/go2/go2_torque,

   python scripts/train.py --task=go2_torque
   

6. Play the trained policy

   python scripts/play.py --task=go2_torque

Troubleshooting:

Contact

• Corresponding author: CAO Yuhong: caoyuhong@nus.edu.sg
• Deployment: LI Peizhuo: lipeizhuo@u.nus.edu
• Policy Learning in Sim: LI Hongyi: hongyi.li@u.nus.edu

Issues

You can create an issue if you meet any bugs, except:

• If you cannot run the vanilla RSL's Legged Gym, it is expected that you first go to the vanilla Legged Gym repo for help.
• There can be CUDA-related errors when there are too many parallel environments on certain PC+GPU+driver combination: we cannot solve thiss, you can try to reduce num_envs.
• Our codebase is only for our hardware system showcased above. We are happy to make it serve as a reference for the community, but we won't tune it for your own robots.

Credit

If our work does help you, please consider citing us and the following works:

@article{li2025sata,
  title={SATA: Safe and Adaptive Torque-Based Locomotion Policies Inspired by Animal Learning},
  author={Li, Peizhuo and Li, Hongyi and Sun, Ge and Cheng, Jin and Yang, Xinrong and Bellegarda, Guillaume and Shafiee, Milad and Cao, Yuhong and Ijspeert, Auke and Sartoretti, Guillaume},
  journal={arXiv preprint arXiv:2502.12674},
  year={2025}
}