1. Installation Guide

November 2, 2023 · View on GitHub

1. Installation Guide

1.1. System Requirement

SimulationFramework has been tested with the following operating systems:
- Ubuntu 18.04, 20.04
- Mac OS
Unsupported systems:
- MS Windows
  - Windows has not been tested by our team.
  - Especially the Mujoco simulation might be lacking on Windows and must be considered highly experimental / unsupported.
  - As a geek in Computer Science, it is time to start your Linux career :D!

1.2. Software Requirements

Before using this repository, you need to install these software packages:
- Mujoco, a state of the art physical simulation engine
  - Mujoco is a physics engine written in C++, which is originally developed at the Movement Control Laboratory, University of Washington and has since been bought (and made open-source) by Google's Deepmind. OpenAI implemented a Python interface (mujoco-py) for this engine, which is utilized by our simulation framework. Since it is tricky to install Mujoco, please follow our install instruction below.
- Anaconda / Miniconda, to handle software dependencies, install and run simulation in a virtual environment.

1.3. Installation Steps

1.3.1. Mujoco

Before starting, you should know:

SimulationFramework can be used with two different versions of Mujoco. If you want the maximum functionality, install both versions of MuJoCo (2.1 and 2.3). Currently, some experiments use MuJoCo 2.1 and some use MuJoCo 2.3.

To install the older MuJoCo version 2.1, you need to follow these instructions:

Download Mujoco 2.1.
- You can download Mujoco from the official Mujoco website. It is important that you install the correct version, please scroll down until you find version 2.1.
Unzip your package and put everything (bin, doc...) under "$HOME/.mujoco/mujoco210/". That's it!

Since Deepmind bought Mujoco, you do not need to acquire a license anymore. It's free!

You should now have the following folder structure in your $HOME:

.mujoco/  
└── mujoco210  
    ├── bin  
    ├── doc  
    ├── include  
    ├── model  
    └── sample

Note that Mujoco is a C++ library. The corresponding Python bindings for MuJoCo 2.1 are called mujoco-py and will be installed by the installation script later. MuJoCo 2.3 will be installed entirely by the installation script, no previous steps are required.

1.3.2 Anaconda / Miniconda

Follow the instructions from the official Installation Guide. Miniconda is sufficient but you can also choose Anaconda.

1.3.3. ALR SimulationFramework (Automatic)

Download / Clone this Git Repository
Install system dependency (because this cannot be installed through conda unfortunately). If you are using lab PCs in the ALR lab, those dependencies should already be installed system-wide, so you can skip this step.
- ```
  sudo apt-get install libosmesa6-dev
```
Automatic Installation:
- In your terminal run the install.sh script to automatically install the ALR SimulationFramework and required dependencies in a new conda environment.
```
bash install.sh
```
Activate your virtual environment (conda activate my_env) and then execute pip install mujoco. This will install MuJoCo version 2.3 as well as its Python bindings.
To allow your PC to communicate with SL running on the PCs that controll the Panda robots, you need to install a module called py_at_broker. You should install it inside your virtual environment, so make sure that it is activated. To install py_at_broker, clone the repository and then execute:
- ```
  cd py_at_broker
  pip install -e python/
```
Ready to go:
- Now your SimulationFramework is ready, you can run a demo script from the demo folder to see if everything works. For example:
  python Demo_Pick_and_Place_IK_Control_Mujoco.py

1.3.4. ALR SimulationFramework Manual Installation Steps

If you for some reason need to install things manually, please refer to our Manual Installation Guide.

1.4. Deinstallation

Since everything is installed in your virtual environment, SimulationFramework will not mess up your global system's dependencies. When you decide to delete Simulation framework, just go with:

conda env remove --name <ENVNAME>

Afterwards delete the downloaded folders (.mujoco/ and the downloaded SimulationFramework)

Back to Overview