Project description

What’s the project about?

Before we start, I think it’s a good idea to answer a question: whats’s the project about? what are we going to talk about in this blog during the next couple of months? The main topic of this project is clear: Autonomous driving using reinforcement learning. We will study about how a robot interacts in a certain enviroment with different algothms and programs and compare them with a reinforcement trained algorithm see how much better reinforment learning actually is and what beneefits and disadvantages does it offer in comparation to other options.

Where are we doing this?

We can break up this questions answer on two diferrent parts different: the simulator and the framework

The simulator

Autonomous driving is something you prabably will end up trying on a real robot. For this simple reason I think it’s an accurate decision to work with a realistic simulator. After some research my choice was to use CARLA as my main simulator, CARLA is a very realistic simulator developed in Barcelona and meant to develop, trainin, and validate autonomous system with a really realistic aproach. I think there couldn’t be any better choice. For more information about carla you can visit their github page: https://github.com/carla-simulator/carla

The Framework

For this point I have chosen one of the best robotic frameworks the exists For this point I have chosen what probabbly one of the best robotic programing frameworks that exists to the date. Im talking about ROS2, ROS2 is my favourite robotics aplications developing framework and one that I have a lot of experience working with. The distribution that we will be using in this particular case will be ROS2 foxy. For more information about ROS2 foxy you can visit their website: https://docs.ros.org/en/foxy/index.html