Introducing PyBullet

PyBullet is a popular high-fidelity physics simulation for robotics, machine learning, games, and more. It is one of the most commonly used libraries for robot learning using RL, especially in sim-to-real transfer research and applications.

Figure 14.1 – PyBullet environments and visualizations (source: PyBullet GitHub repo)

PyBullet allows developers to create their own physics simulations. In addition, it has prebuilt environments using the OpenAI Gym interface. Some of those environments are shown in Figure 14.1.

In the next section, we will set up a virtual environment for PyBullet.

Setting up PyBullet

It is almost always a good idea to work in virtual environments for Python projects, which is also what we do for our robot learning experiments in this chapter. So, let's go ahead and execute the following commands to install the libraries we will use:

$ virtualenv pybenv
$ source pybenv/bin/activate
$ pip install...