RoMoCo

RoMoCo (Reduced-order Modeling and Control) is a C++ framework for simulation and control of bipedal and humanoid robots. It provides modular components for planning, control, and simulation interfaces, with the goal of offering a unified, extensible framework for research and development.

Features

• Modular design for walking, standing, and in-air robot control
• Pinocchio-based kinematics and dynamics
• Integration with MuJoCo for simulation
• Support for state-of-the-art reduced-order model planners (Alip, H-LIP, MLIP flat footed, DCM)
• Multiple whole-body control methods: TSCQP, position IK, velocity IK, inverse dynamics
• Easy-to-extend base classes for new robots or controllers

Overview

• romoco_core: Core functionalities: Data structures and types used throughout the control framework, robot kinematics, dynamics, and base classes.
• romoco_planner: Reduced-order model planners
• romoco_output: Output generation and embedding for whole-body control
• romoco_control: Different whole-body torque control methods
• romoco_mujoco: Low-level simulation interface for MuJoCo
• romoco_state_machine: Example state machine for bipedal locomotion
• romoco_utils: Utility functions: Bezier, Yaml parsing, geometry, etc.
• romoco_msgs: ROS 2 message definitions
• romoco_screen_radio: a on screen radio interface for robot control

Robots: G1, Cassie, H1, more to come

Build & Run

Tested with ROS 2 Humble and Ubuntu 22.04 (Jammy).

Documentation

Doxygen documentation can be generated by running:

doxygen docs/Doxyfile

Then open docs/html/index.html in a web browser.

Run from Docker

./docker/build.sh
./docker/run.sh

Build locally

Install the dependencies listed below then clone the repository and build it using colcon build.

colcon build --symlink-install --cmake-args -DCMAKE_BUILD_TYPE=Release
. install/setup.bash

Build

Install the dependencies listed below then clone the repository and build it using colcon build.

Run Tests

colcon test

Launch

• ROS 2 (recommended):
  • Source the workspace: . install/setup.bash
  • Example launch commands (adjust names if different):
    • ros2 launch g1_stack g1mujoco.launch.xml or ros2 launch g1_stack g1mujoco.launch.py
    • ros2 launch cassie_stack cassiemujoco.launch.py (includes a fake radio; toggle SB from 0→1 for stand-to-walk)

License

This project is licensed under the GNU General Public License v3.0. See the LICENSE file for details.

Installation

sudo apt-get update
sudo apt-get install -y \
  build-essential vim gcc g++ gfortran git patch wget cmake \
  python3-pip liblapack-dev libmetis-dev libblas-dev libatlas-base-dev \
  libglpk-dev pkg-config ipython3 python3-dev python3-tk swig \
  doxygen doxygen-latex liburdfdom-dev libassimp-dev libboost-all-dev \
  libglfw3 libglfw3-dev qtcreator qtbase5-dev lsb-release --install-recommends

python3 -m pip install --upgrade pip numpy

echo 'export PATH=$HOME/.local/bin:$PATH' >> ~/.bashrc
source ~/.bashrc

EigenPy

Repository: https://github.com/stack-of-tasks/eigenpy/tree/master

On Ubuntu 22.04 (Python 3.10):

sudo sh -c "echo 'deb [arch=amd64] http://robotpkg.openrobots.org/packages/debian/pub $(lsb_release -cs) robotpkg' >> /etc/apt/sources.list.d/robotpkg.list"
curl http://robotpkg.openrobots.org/packages/debian/robotpkg.key | sudo apt-key add -
sudo apt-get update
sudo apt install robotpkg-py310-eigenpy

Also add its path to your .bashrc:

echo 'export CMAKE_PREFIX_PATH=/opt/openrobots:$CMAKE_PREFIX_PATH' >> ~/.bashrc

If your Python minor version differs, install the matching package (e.g., robotpkg-py39-eigenpy for Python 3.9).

Pinocchio

mkdir -p ~/repos && cd ~/repos
git clone --recursive https://github.com/stack-of-tasks/pinocchio
cd pinocchio
mkdir build && cd build
cmake .. \
  -DCMAKE_BUILD_TYPE=Release \
  -DCMAKE_INSTALL_PREFIX=/usr/local \
  -DCMAKE_PREFIX_PATH="/opt/openrobots"
make -j4
sudo make install

Update your .bashrc (PYTHONPATH adapted dynamically to your Python version):

export PATH=/usr/local/bin:$PATH
export PKG_CONFIG_PATH=/usr/local/lib/pkgconfig:$PKG_CONFIG_PATH
export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH
export PYTHONPATH=/usr/local/lib/python3.10/site-packages:$PYTHONPATH
export CMAKE_PREFIX_PATH=/usr/local:$CMAKE_PREFIX_PATH

Eigen 3.4.0

cd ~/repos/
git clone https://gitlab.com/libeigen/eigen.git 
cd eigen 
git checkout 3147391d 
mkdir build 
cd build 
cmake .. 
sudo make install

MuJoCo

mkdir -p ~/repos && cd ~/repos
wget https://github.com/google-deepmind/mujoco/releases/download/3.2.6/mujoco-3.2.6-linux-x86_64.tar.gz
tar -xvzf mujoco-3.2.6-linux-x86_64.tar.gz
rm mujoco-3.2.6-linux-x86_64.tar.gz 


### QP (Clarabel.cpp)
```bash
mkdir -p ~/repos && cd ~/repos
git clone --recurse-submodules https://github.com/oxfordcontrol/Clarabel.cpp.git
cd Clarabel.cpp
mkdir build && cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
cmake --build . 

If installing to a non-standard prefix, add it to CMAKE_PREFIX_PATH.

For debugging (Eigen pretty printers)

mkdir -p ~/repos && cd ~/repos
git clone https://github.com/dmillard/eigengdb
cd eigengdb
sudo python3 setup.py install
python3 bin/eigengdb_register_printers

Todo

• Hierarchical QP controller for humanoids
• Multi-domain walking

G1 Interface

Network setup and instructions for controlling the G1 robot through DDS can be found here: https://support.unitree.com/home/en/G1_developer

Cassie Interface

Calibration, development PC, and network setup instructions for controlling the Cassie robot through UDP and Simulink Real-Time Explorer target PC can be found here: https://github.com/agilityrobotics/cassie-doc/wiki