Abstract: In this article, we present a hierarchical supervisory reinforcement learning (RL) framework to achieve precise trajectory tracking for UAV(s) operating in dynamic and complex environments.