Improve an existing Reinforcement Learning project

Hi, I propose that we discuss what you need to acheive with the tool decribed in the medium article. We can then proceed to implement and train the steering network. My current understanding is that this is required to develop the new features in your project.