Deep Learning, also known as deeply structured or hierarchical learning, is part of a larger family of machine learning methods. DL is applied to robin and roy in a controlled manner. Controlled DL means that the algorithm gets examples of input and corresponding output. Based on these examples it learns how the characteristics of the input define the output. After the learning phase, the algorithm will be able to produce the right output for new input. The results are very similar to the results of human hands – and often even better!