Deep Driving Workshop

On July 10, a practical workshop on machine learning and autonomous driving was held for 11th grade students at Archenhold-Gymnasium in Berlin. The event began with a brief introduction to neural networks and their applications in machine learning. The students then put these concepts into practice by working with an autonomous model vehicle. Their task was to remotely control the vehicle and collect extensive LiDAR sensor data and driving maneuver data to train a neural network.

After training the neural network, the students tested its performance by observing how well it could handle driving tasks based on the learned maneuvers. To challenge the neural network, the students changed them during the testing phase and introduced new scenarios that the network was not yet familiar with. By iteratively testing and adjusting the neural network's hyperparameters, the students were able to refine its performance and ultimately improve the vehicle's ability to navigate autonomously.