Abstract: In this paper we combine Reinforcement Learning (RL), Behavior Trees (BTs), and Control Barrier Functions (CBFs) to create controllers for complex tasks. Pairwise, these tools have been ...