A Deep Reinforcement Learning Framework (simulation) with OpenAI-Gym and PyBullet Environment for the 7-DOF manipulator downscale version from Robotory - SKKU

Prerequisites:

• OpenAI Gym: https://github.com/openai/gym
• PyBullet: https://github.com/bulletphysics/bullet3
• Virtualenv or Anaconda: https://phoenixnap.com/kb/how-to-install-anaconda-ubuntu-18-04-or-20-04

I) OpenAI Gym Environment for SKKU Robotory 7-DOF manipulator downscale version

There are 2 method for installing our Deep RL OpenAI Gym environment

1) Install directly:

• Download gym-robotorydownscale
• Move to this folder and:

pip3 install gym-robotorydownscale

2) Building from source:

git clone https://github.com/shinhoang88/gym-robotorydownscale.git
cd gym-robotorydownscale
pip3 install -e .

II) Download the 7-DOF manipulator URDF description:

1) Source:

git clone https://github.com/shinhoang88/downscale.git

2) Copy it into your virtualenv or anaconda3 python/pybullet_data workspace, for example:

sudo cp -r /home/username/downscale/ /home/username/rospython3_ws/rospy3env/lib/python3.6/site-packages/pybullet_data/

III) Test the environment working:

import gym
import gym_robotorydownscale
import pybullet as p
import pybullet_data

# Initialize the OpenAI Gym environment
env = gym.make('robotorydownscale-v0')
for i_episode in range(20):
    # Reset the environment
    observation = env.reset()
    for t in range(18000):
        # Stochastic action sample inside the action_space box
        action = env.action_space.sample()
        # Getting info from environment step simulation
        observation, reward, done, info = env.step(action)
        if done:
            print("Episode finished after {} timesteps".format(t + 1))
            break
env.close()

• Author : Phi Tien Hoang
• E-mail : [email protected]
• Organization : Robotory-SKKU-S.Korea