Reinforcement learning for FTC. This notebook attempts to get a robot to travel to waypoints(rings) with a basic MLP. The network has an observation space of robot pose, pose velocity, information about the target pose, and the timestep. The action space is forwards, sideways, and rotational acceleration.
