Course

Using OpenAI with ROS Python

Use the power of OpenAI combined with ROS simulations in the easiest way

ROS Noetic

Course Overview

Description

In this Course, you are going to learn how to use the OpenAI ROS structure developed by The Construct and how to generate new code for it. The OpenAI ROS structure will allow you to develop for OpenAI with ROS in a much easier way.

Learning Objectives

  • Understand the basic concepts of the OpenAI ROS structure
  • Set up the OpenAI ROS structure for a CartPole environment
  • Train the Cartpole with the Qlearn algorithm
  • Set up the OpenAI ROS structure for a Moving Cube environment
  • Train the Cube with the Qlearn algorithm
  • Modify the learning algorithm: DeepQ
  • Set up the OpenAI ROS structure for a Fetch Robot 
  • Train a Fetch robot with the HER algorithm from OpenAI baselines

Simulation robots used in this course

Fetch Robot, CartPole, Hopper, RoboCube. 

Level

Advanced

}

Duration

25h 30m

Prerequisites

This course is part of this learning path:

What projects will you be doing?

[ROS Q&A] 168 - What are the differences between global and local costmap

Training a Fetch Robot

Build the Task Environment for training a Fetch robot

ROS Mini Challenge #2 - RViz

Training a Hopper robot

Create environments to train the one-legged Hopper robot

OpenAI-with-Moving-Cube-Robot-in-Gazebo-Step-by-Step-Part2

Apply openai_ros to a robot

Apply the openai_ros package to a cube robot

Using the CartPole simulated environment

Understanding the ROS + OpenAI structure

What you will learn

Course Syllabus

Unit 1: Introduction

A brief introduction to the course contents. Includes a demo.

10 min.

Unit 2: Exploring the OpenAI Structure: CartPole

Explore the OpenAI structure, using a CartPole environment.

4 hrs.

Unit 3: Exploring the OpenAI Structure: RoboCube. Part 1

Explore the OpenAI structure, using a RoboCube simulation. Presentation.

10 min.

Unit 4: Exploring the OpenAI Structure: RoboCube. Part 2

Explore the OpenAI structure, using a RoboCube simulation. Create the Robot Environment.

2 hrs. 

Unit 5: Exploring the OpenAI Structure: RoboCube. Part 3

Explore the OpenAI structure, using a RoboCube simulation. Create the Task Environment.

2 hrs. 

Unit 6: Save and Load the Learned Policy

Learn how to save and load the learned policy

1 hr. 

Unit 7: Modifying the Learning algorithm: CartPole

How to modify the learning algorithm for the CartPole environment. We’ll use the deepQ algorithm.

2 hrs. 

Unit 8: Modifying the Learning algorithm: RoboCube

How to modify the learning algorithm for the RoboCube environment. We’ll use the deepQ algorithm.

2 hrs. 

Unit 9: Training a Fetch Robot. Part 1

Create a Robot Environment for the Fetch robot.

2 hrs. 

Unit 10: Training a Fetch Robot. Part 2

Create a Task Environment for the Fetch robot. We’ll use the HER algorithm for training Fetch robot.

2 hrs. 

Unit 11: Project: Training a Hopper robot

Apply everything you have learned for training a Hopper robot.

8 hrs. 

Ready to get started?

Start learning ROS & Robotics online quickly and easily

What’s next

People interested in this course also viewed

Course

ROS Deep Learning with TensorFlow 101

Course

Deep Learning with Domain Randomization

Top universities choose The Construct for Campus to teach ROS & Robotics.

Pin It on Pinterest