RAMMP

RAMMP can be deployed in one of the following configurations:

• Two-machine setup
  • GPU Machine (GM): Runs compute-intensive components.
  • Controller Machine (CM): Handles controllers and emergency stops.
• Single-machine setup (SM): Runs all components on a single system.

---

Requirements

• Operating System: Ubuntu 20.04
• Python: 3.10 or later
• ROS: Noetic (including rospy)

---

Installation

The system uses two Conda environments:

• compute – for compute-intensive components
• controller – for control and hardware interfaces

In a two-machine setup, these environments typically live on separate machines (GM and CM).
In a single-machine setup, both environments can exist on the same machine.

---

Compute Setup

1. Create the Conda environment:

   conda create -n compute python=3.10

2. Activate the environment:

   conda activate compute

3. Install PyTorch (CUDA 12.1):

   conda install pytorch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 \
       pytorch-cuda=12.1 numpy=1.26.4 numpy-base=1.26.4 \
       -c pytorch -c nvidia -c defaults

4. Install PyTorch3D:

   python -m pip install pytorch3d \
       -f https://dl.fbaipublicfiles.com/pytorch3d/packaging/wheels/py310_cu121_pyt210/download.html

5. Clone the RAMMP repository:

   git clone https://github.com/empriselab/RAMMP
   cd RAMMP

6. Install additional dependencies:

   pip install --no-build-isolation chumpy

7. Install RAMMP (full installation):

   pip install -e ".[full]"

8. Install Pinocchio:

   conda install -c conda-forge "pinocchio=3.1.*"

9. Add the following to your ~/.bashrc:

   export LD_LIBRARY_PATH=/lib/x86_64-linux-gnu:/usr/lib/x86_64-linux-gnu:/lib:/usr/local/cuda-12.1/lib64:/opt/ros/noetic/lib:/opt/ros/noetic/lib/x86_64-linux-gnu
   export LD_PRELOAD="/lib/x86_64-linux-gnu/libffi.so.7 /lib/x86_64-linux-gnu/libtiff.so.5"

   Then reload:

   source ~/.bashrc

---

Controller Setup

1. Create the Conda environment:

   conda create -n controller python=3.10

2. Activate the environment:

   conda activate controller

3. Install the Kortex API (after downloading the wheel file):

   python -m pip install kortex_api-2.6.0.post3-py3-none-any.whl

4. Clone the RAMMP repository:

   git clone https://github.com/empriselab/RAMMP
   cd RAMMP

5. Install RAMMP (controller-only installation):

   pip install -e ".[controller]"

6. Install Pinocchio:

   conda install -c conda-forge "pinocchio=3.1.*"

---

Running

Work in progress.