diff --git a/.github/workflows/docker.yaml b/.github/workflows/docker.yaml new file mode 100644 index 0000000..6caf658 --- /dev/null +++ b/.github/workflows/docker.yaml @@ -0,0 +1,17 @@ +name: Publish Docker Image + +on: [workflow_dispatch] + +jobs: + publish: + runs-on: ubuntu-latest + steps: + - uses: actions/checkout@v4 + - name: Publish to Registry + uses: elgohr/Publish-Docker-Github-Action@v5 + with: + name: ami-iit/dnn-mpc-walking-docker + username: ${{ github.actor }} + password: ${{ secrets.GITHUB_TOKEN }} + workdir: dockerfiles + registry: ghcr.io diff --git a/CMakeLists.txt b/CMakeLists.txt index 69c2ae3..53e8732 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -6,7 +6,7 @@ cmake_minimum_required(VERSION 3.14) ## MAIN project project(DNNMPCWalking - VERSION 0.0.1) + VERSION 1.0.0) list(APPEND CMAKE_MODULE_PATH ${CMAKE_CURRENT_SOURCE_DIR}/cmake) diff --git a/README.md b/README.md index 99b04d6..6724bff 100644 --- a/README.md +++ b/README.md @@ -3,126 +3,131 @@ Online DNN-Driven Nonlinear MPC for Stylistic Humanoid Robot Walking with Step A
-Giulio Romualdi, Paolo Maria Viceconte, Lorenzo Moretti, Ines Sorrentino, Stefano Dafarra, Silvio Traversaro, and Daniele Pucci -
-Paolo Maria Viceconte and Giulio Romualdi are co-first authors +Giulio Romualdi, Paolo Maria Viceconte, Lorenzo Moretti, Ines Sorrentino, Stefano Dafarra, Silvio Traversaro, and Daniele Pucci +
+Co-first authors: Paolo Maria Viceconte and Giulio Romualdi

-📅 This paper has been accepted for publication at the 2024 IEEE-RAS International Conference on Humanoid Robots (Humanoids), Nancy, France 🤖 +📅 Accepted for publication at the 2024 IEEE-RAS International Conference on Humanoid Robots (Humanoids), Nancy, France 🤖

- 📚 Paper     - 🎥 Video     - 🖼️ Poster     - 🔧 Experiments     - 🌐 Website     - 📂 Dataset + 📚 Paper     + 🎥 Video     + 🖼️ Poster     + 🔧 Experiments     + 🌐 Website     + 📂 Dataset

-https://github.com/user-attachments/assets/404c9af8-528e-43c2-abd2-138a98adfc04 - -
+--- ## Reproducing the Experiments -You can reprocue the experiments in two ways: either with conda or pixi. - -### Conda - +You can reproduce the experiments using **Docker**, **Conda**, or **Pixi**. -To reproduce the experiments, we provide a conda environment for easy setup. Follow the steps below: +### Docker -#### 1. Install the Environment -Run the following command to create the conda environment: +Run the experiments via Docker for an isolated and reproducible environment. -```bash -conda env create -f environment.yml -``` +1. Pull the Docker image: + ```bash + docker pull ghcr.io/ami-iit/dnn-mpc-walking-docker:latest + ``` -#### 2. Activate the Environment -Activate the newly created environment: +2. Launch the container: + ```bash + xhost + + docker run -it --rm \ + --device=/dev/dri:/dev/dri \ + --env="DISPLAY=$DISPLAY" \ + --net=host \ + ghcr.io/ami-iit/dnn-mpc-walking-docker:latest + ``` -```bash -conda activate dnn-mpc-env -``` +3. Wait for `Gazebo` to start and launch the experiment. -#### 3. Run the Simulation -Start the experiment with: +> ⚠️ **Known Issue:** The Gazebo real-time factor is scaled by a factor of 10 due to the MUMPS linear solver in the IPOPT Docker image. Alternative solvers (e.g., MA97) are available but cannot be redistributed. -```bash -./run_simulation.sh -``` - -This script will: -- Launch the Gazebo simulator -- Start the YARP server (for simulator communication) -- Initialize the DNN-driven MPC controller +--- -When prompted, type `y` and press `Enter` to start the simulation. The humanoid robot will begin walking, and you can observe its behavior in the Gazebo simulator. +### Conda -⚠️ **Known Issue:** The Gazebo real-time factor is scaled by a factor of 10. This is due to the use of the MUMPS linear solver in the IPOPT docker image. Alternative solvers (e.g., MA27) are available but cannot be redistributed. +Follow these steps to set up the experiments using Conda: +1. Install the environment: + ```bash + conda env create -f environment.yml + ``` -### Pixi +2. Activate the environment: + ```bash + conda activate dnn-mpc-env + ``` -To run the experiments with pixi on Linux, just download the repo and run: +3. Compile the code: + ```bash + cd paper_romualdi_viceconte_2024_humanoids_dnn-mpc-walking + mkdir build && cd build + cmake .. + make -j + make install + ``` -~~~ -git clone https://github.com/ami-iit/paper_romualdi_viceconte_2024_humanoids_dnn-mpc-walking -cd paper_romualdi_viceconte_2024_humanoids_dnn-mpc-walking -pixi run -e default run_simulation -~~~ +4. Run the simulation: + ```bash + ./run_simulation.sh + ``` -This command will install all the dependencies, compiled the code and run the simulation. At that point, when prompted, type `y` and press `Enter` to start the simulation. The humanoid robot will begin walking, and you can observe its behavior in the Gazebo simulator. +> ⚠️ The Gazebo real-time factor is scaled by a factor of 10 due to the MUMPS linear solver. -⚠️ **Known Issue:** The Gazebo real-time factor is scaled by a factor of 10. This is due to the use of the MUMPS linear solver in the IPOPT docker image. If you have have access to a Coin-HSL license, you can use it following the instructions in the following, to reduce the Gazebo real-time factor scaling from 10 to 2. - -To run the simulation using the Coin-HSL's `ma97` solver, follow the following steps: +--- -1. Go to https://licences.stfc.ac.uk/product/coin-hsl, and: - - If you already have a license for Coin-HSL: - - Go to https://licences.stfc.ac.uk/account/orders and find the coin-hsl order. - - Download the coinhsl-2023.11.17.zip file and place it in the './coinhsl_src' folder of this repository. - - If you do not have a license for Coin-HSL: - - If you are an academic, request a license at https://licences.stfc.ac.uk/product/coin-hsl. - - If you are not an academic, purchase a license at https://licences.stfc.ac.uk/product/coin-hsl. - - Once your order is approved, download the coinhsl-2023.11.17.zip file and place it in the './coinhsl_src' folder. +### Pixi -Once the `coinhsl-2023.11.17.zip` archive is in the './coinhsl_src' folder, just run: +To run the experiments with Pixi: -~~~ -pixi run -e coinhsl run_simulation -~~~ +1. Clone the repository: + ```bash + git clone https://github.com/ami-iit/paper_romualdi_viceconte_2024_humanoids_dnn-mpc-walking + cd paper_romualdi_viceconte_2024_humanoids_dnn-mpc-walking + ``` -This will execute the same steps of running `pixi run -e default run_simulation`, but additionally it will: -* compile `coinhsl` to be able to use the `ma97` linear solver, -* it will modify the configuration files to use `ma97` -* use a different Gazebo world model, to ensure a faster simulation. +2. Run the simulation: + ```bash + pixi run -e default run_simulation + ``` +> **Using MA97 Solver (Optional):** +> If you have access to the Coin-HSL license, you can use the MA97 solver to improve performance: +> 1. Obtain the Coin-HSL archive (`coinhsl-2023.11.17.zip`) and place it in the `./coinhsl_src` folder. +> 2. Run: +> ```bash +> pixi run -e coinhsl run_simulation +> ``` --- ## Maintainers - - - - - +
- - Paolo Maria Viceconte
- 👨‍💻 @paolo-viceconte - - 		- - Giulio Romualdi
- 👨‍💻 @GiulioRomualdi - -
+ + + +
+ + Paolo Maria Viceconte
+ 👨‍💻 Paolo Maria Viceconte + + 		+ + Giulio Romualdi
+ 👨‍💻 Giulio Romualdi + +
diff --git a/dockerfiles/Dockerfile b/dockerfiles/Dockerfile index 17a9681..de8f1e7 100644 --- a/dockerfiles/Dockerfile +++ b/dockerfiles/Dockerfile @@ -35,6 +35,7 @@ WORKDIR /workspace # Clone the repository with the submodules compile it and install it in the conda environment RUN git clone https://github.com/ami-iit/paper_romualdi_viceconte_2024_humanoids_dnn-mpc-walking.git && \ cd paper_romualdi_viceconte_2024_humanoids_dnn-mpc-walking && \ + git checkout v1.0.0 && \ mkdir build && \ cd build && \ cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/opt/conda/envs/dnn-mpc-env .. && \