Awesome

MEVIUS

This repository includes the hardware and software components of "MEVIUS: A Quadruped Robot Easily Constructed through E-Commerce with Sheet Metal Welding and Machining."

Please refer to the following links for more information.

Hardware

You can find all the hardware components in the following link

Goodle Drive

From the link of RUBIK LINK V2.0, please download the upper computer installation program and make all the motors MIT mode.

AK70-10
AK Serials Manual
RUBIK LINK V2.0
Motor Order: FR(1-3) --> FL(4-6) --> BR(7-9) --> BL(10-12) / Proximal --> Distal
Link Name: FR_scapula_link --> FR_thigh_link --> FR_calf_link
Joint Name: FR_collar --> FR_hip --> FR_knee

Software Dependencies

Please install the following dependencies.

Please install the following python packages.

bitstring
pytorch
scipy
mujoco
mujoco-python-viewer

Build

mkdir -p ~/mevius_ws/src
cd ~/mevius_ws/src
git clone git@github.com:haraduka/mevius.git
cd ..
catkin build
source ~/mevius_ws/devel/setup.bash

Reinforcement Learning

Caution: This policy is very simple and not so stable. Currently, it became very stable with a lot of improvement and will be uploaded soon.

# in legged_gym
git remote add haraduka git@github.com:haraduka/legged_gym.git
git fetch haraduka
git checkout haraduka/mevius

# train with two phases
# first phase
python3 legged_gym/scripts/train.py --task mevius
# second phase
# change the parameters in legged_gym/envs/mevius/mevius_config.py from "first" to "second"
# there are two parameters to change: commands.ranges and domain_rand.friction_range
python3 legged_gym/scripts/train.py --task mevius
python3 legged_gym/scripts/play.py --task mevius --load_run ( log_dir )
cp logs/flat_mevius/( log_dir )/exported/policies/policy_1.pt ../mevius/models/policy.pt

Usage

Please test the installation by running the following command.

python3 scripts/mevius_utils.py

Please test the motors by running the following command.

./bin/can_setup.sh
python scripts/tmotor_test.py --ids 1 --task sense

Please test the real robot by running the following command.

# in the PC of MEVIUS
roscore
roslaunch realsense2_camera rs_t265.launch
./bin/can_setup.sh
python3 scripts/mevius_main.py

# in your local PC
roslaunch spacenav_node classic.launch

# Left Click: SITDOWN/STANDUP, Right Click: STANDUP/WALK

# if you want to simulate the robot in mujoco
python3 scripts/mevius_main.py --sim

# please press D to turn off every-step rendering in mujoco viewer

Acknowledgements

scripts/tmotor_lib.py is from mini-cheetah-tmotor-python-can
scripts/isaacgym_torch_utils.py is from IsaacGym
scripts/legged_gym_math.py is from LeggedGym

Citation

@inproceedings{kawaharazuka2024mevius,
  author={K. Kawaharazuka and S. Inoue and T. Suzuki and S. Yuzai and S. Sawaguchi and K. Okada and M. Inaba},
  title={{MEVIUS: A Quadruped Robot Easily Constructed through E-Commerce with Sheet Metal Welding and Machining}},
  booktitle={Proceedings of the 2024 IEEE-RAS International Conference on Humanoid Robots},
  year=2024,
}