This report aims to demonstrate how Reinforcement Learning can be used in real world tasks, build a robust model and have the ability to adapt to uncertainties. The report will follow simulations and programming that aims to demonstrate how one could speed up and simplify calculations needed for real world problems though the use of Reinforcement Learnt neural networks.

This will be demonstrated through a project implementation which will be simulating a Mars landing. The simulation will also implement a Terrain Relative Navigation tool used to demonstrate how Reinforcement Learning can be supported by other algorithms and produce a more robust mode.