Built and trained a model that determines the steering angle the car needs to take to round the track. The open-source simulator has a training mode and an autonomous mode. In training mode, the user records their driving around the track. Photos from a left, center and right camera through the driver’s perspective are written to file along with the speed and steering angle. That’s enough data to train a convolutional neural network. With a trained CNN, autonomous mode can be used to see how the model performs on the track.