Reinforcement Learning (RL) has shown remarkable success in real-world applications, particularly in robotics control. However, RL adoption remains limited due to insufficient safety guarantees. We introduce Nightmare Dreamer, a model-based Safe RL algorithm that addresses safety concerns by leveraging a learned world model to predict potential safety violations and plan actions accordingly. Nightmare Dreamer achieves nearly zero safety violations while maximizing rewards. Nightmare Dreamer outperforms model-free baselines on Safety Gymnasium tasks using only image observations, achieving nearly a 20x improvement in efficiency.
@article{arxiv.2601.04686,
title = {Nightmare Dreamer: Dreaming About Unsafe States And Planning Ahead},
author = {Oluwatosin Oseni and Shengjie Wang and Jun Zhu and Micah Corah},
journal= {arXiv preprint arXiv:2601.04686},
year = {2026}
}
Comments
RSS'25: Multi-Objective Optimization and Planning in Robotics Workshop: 5 pages, 8 figures