NEWS

2017 July 17, Version 1.1

• All envs version bumped to “-v1", due to stronger stuck joint punishment, that improves odds of getting a good policy.
• Flagrun-v1 is much more likely to develop a symmetric gait,
• FlagrunHarder-v1 has new "repeat-underlearned" learning schedule, that allows it to be trained to stand up, walk and turn without falling.
• Atlas robot model, modified (empty links removed, overly powerful feet weakaned).
• All -v1 envs are shipped with better zoo policies, compared to May versions.
• Keyboard-controlled humanoid included.

Roboschool

Release blog post is here:

https://blog.openai.com/roboschool/

Roboschool is a long-term project to create simulations useful for research. The roadmap is as follows:

1. Replicate Gym MuJoCo environments.
2. Take a step away from trajectory-centric fragile MuJoCo tasks.
3. Explore multiplayer games.
4. Create tasks with camera RGB image and joints in a tuple.
5. Teach robots to follow commands, including verbal commands.

Some wiki pages:

Contributing New Environments

Help Wanted

Environments List

The list of Roboschool environments is as follows:

• RoboschoolInvertedPendulum-v0
• RoboschoolInvertedPendulumSwingup-v0
• RoboschoolInvertedDoublePendulum-v0
• RoboschoolReacher-v0
• RoboschoolHopper-v0
• RoboschoolWalker2d-v0
• RoboschoolHalfCheetah-v0
• RoboschoolAnt-v0
• RoboschoolHumanoid-v0
• RoboschoolHumanoidFlagrun-v0
• RoboschoolHumanoidFlagrunHarder-v0
• RoboschoolPong-v0

To obtain this list: import roboschool, gym; print("\n".join(['- ' + spec.id for spec in gym.envs.registry.all() if spec.id.startswith('Roboschool')])).

Installation

First, define a ROBOSCHOOL_PATH variable in the current shell. It will be used in this README but not anywhere in the Roboschool code.

ROBOSCHOOL_PATH=/path/to/roboschool

If you have both Python2 and Python3 on your system, use python3 and pip3 commands.

The dependencies are gym, Qt5, assimp, tinyxml, and bullet (from a branch). For the non-bullet deps, there are several options, depending on what platform and package manager you are using.

• Ubuntu:

  apt install cmake ffmpeg pkg-config qtbase5-dev libqt5opengl5-dev libassimp-dev libpython3.5-dev libboost-python-dev libtinyxml-dev

  Users report in issue #15 that sudo pip3 install pyopengl can make OpenGL errors go away, because it arranges OpenGL libraries in an Ubuntu system in the right way.

• Linuxbrew

  brew install boost-python --without-python --with-python3 --build-from-source
  export C_INCLUDE_PATH=/home/user/.linuxbrew/include:/home/user/.linuxbrew/include/python3.6m
  export CPLUS_INCLUDE_PATH=/home/user/.linuxbrew/include:/home/user/.linuxbrew/include/python3.6m
  export LIBRARY_PATH=/home/user/.linuxbrew/lib
  export PKG_CONFIG_PATH=/home/user/.linuxbrew/lib/pkgconfig:/usr/lib/pkgconfig:/usr/lib/x86_64-linux-gnu/pkgconfig

  (still use Qt from Ubuntu, because it's known to work)

• Mac, homebrew python:

  # Will not work on Mavericks: unsupported by homebrew, some libraries won't compile, upgrade first
  brew install python3
  brew install cmake tinyxml assimp ffmpeg qt
  brew reinstall boost-python --without-python --with-python3 --build-from-source
  export PATH=/usr/local/bin:/usr/local/opt/qt5/bin:$PATH
  export PKG_CONFIG_PATH=/usr/local/opt/qt5/lib/pkgconfig

• Mac, Anaconda with Python 3

  brew install cmake tinyxml assimp ffmpeg
  brew reinstall boost-python --without-python --with-python3 --build-from-source
  conda install qt
  export PKG_CONFIG_PATH=$(dirname $(dirname $(which python)))/lib/pkgconfig

Compile and install bullet as follows. Note that make install will merely copy files into the roboschool directory.

git clone https://github.com/olegklimov/bullet3 -b roboschool_self_collision
mkdir bullet3/build
cd    bullet3/build
cmake -DBUILD_SHARED_LIBS=ON -DUSE_DOUBLE_PRECISION=1 -DCMAKE_INSTALL_PREFIX:PATH=$ROBOSCHOOL_PATH/roboschool/cpp-household/bullet_local_install -DBUILD_CPU_DEMOS=OFF -DBUILD_BULLET2_DEMOS=OFF -DBUILD_EXTRAS=OFF  -DBUILD_UNIT_TESTS=OFF -DBUILD_CLSOCKET=OFF -DBUILD_ENET=OFF -DBUILD_OPENGL3_DEMOS=OFF ..
make -j4
make install
cd ../..

Finally, install project itself:

pip3 install -e $ROBOSCHOOL_PATH

Now, check to see if it worked by running a pretrained agent from the agent zoo.

Agent Zoo

We have provided a number of pre-trained agents in the agent_zoo directory.

To see a humanoid run towards a random varying target:

python $ROBOSCHOOL_PATH/agent_zoo/RoboschoolHumanoidFlagrun_v0_2017may.py

To see three agents in a race:

python $ROBOSCHOOL_PATH/agent_zoo/demo_race2.py