To Use Deep Learning to predict the current location of a Reinforcement learning agent based on point cloud (3D) data. A combination of CNN and RNN is used to determine the X, Y, and Z coordinates and an angle theta, which denotes the angle at which the object is moving.