Exploring Safe Reinforcement Learning for Sequential Decision Making - Robotics Institute Carnegie Mellon University