Description

Force-feedback MPC based on online estimation. This repo implements custom residual, action model and estimator in C++ (with Python bindings) based on the Crocoddyl library. This is meant to be used as a plugin to reproduce the work described in this publication.

Dependencies

Core (C++/Python bindings)

• Crocoddyl
• Pinocchio
• Proxsuite

Demos (Python)

• croco_mpc_utils
• mim_robots
• PyBullet
• matplotlib
• PyYAML

Install from source

Setup environment

You can optionally use conda to setup your work environment

conda create -n force_observer

conda activate force_observer

conda install -c conda-forge mim-solvers cmake proxsuite

conda install conda-forge::pyyaml

conda install matplotlib

conda install conda-forge::pybullet

Also check out environment.yaml to full conda environment. Then you can install manually the remaining machines-in-motion dependencies croco_mpc_utils and mim_robots (use the pip install . --no-deps).

Build and install

Then clone and build / install the code

git clone [email protected]:machines-in-motion/force_observer.git

git submodule update --init

mkdir build && cd build

cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX={INSTALL_DIR}

make && sudo make install

To install inside the conda environment, you can use -DCMAKE_INSTALL_PREFIX=$CONDA_PREFIX (environment must be activated).

How to use it

In demos run the contact or sanding task script, e.g. python sanding_mpc.py. You can modify the corresponding config file, e.g. sanding_mpc.yml.

Run the unit test from the build folder by running ctest -v

Import the python bindings of C++ classes with import force_observer

Citing this repo

@INPROCEEDINGS{10611156,
  author={Jordana, Armand and Kleff, Sébastien and Carpentier, Justin and Mansard, Nicolas and Righetti, Ludovic},
  booktitle={2024 IEEE International Conference on Robotics and Automation (ICRA)}, 
  title={Force Feedback Model-Predictive Control via Online Estimation}, 
  year={2024},
  volume={},
  number={},
  pages={11503-11509},
  keywords={Torque;Systematics;Force;Force feedback;Estimation;Robot sensing systems;Force sensors},
  doi={10.1109/ICRA57147.2024.10611156}}

A. Jordana, S. Kleff, J. Carpentier, N. Mansard and L. Righetti, "Force Feedback Model-Predictive Control via Online Estimation," 2024 IEEE International Conference on Robotics and Automation (ICRA), Yokohama, Japan, 2024, pp. 11503-11509, doi: 10.1109/ICRA57147.2024.10611156. keywords: {Torque;Systematics;Force;Force feedback;Estimation;Robot sensing systems;Force sensors},