Abstract: This article introduces a sustainable human-guided reinforcement learning (RL) framework to address the challenge of learning performance degradation when the human guidance is suspended.