Real-DRL: Teach and Learn in Reality
On the map: Deep reinforcement learning methods → Safe and constrained RL
Abstract (arXiv)
This paper introduces the Real-DRL framework for safety-critical autonomous systems, enabling runtime learning of a deep reinforcement learning (DRL) agent to develop safe and high-performance action policies in real plants (i.e., real physical systems to be controlled), while prioritizing safety! The Real-DRL consists of three interactive components: a DRL-Student, a PHY-Teacher, and a Trigger. The DRL-Student is a DRL agent that innovates in the dual self-learning and teaching-to-learn paradigm and the real-time safety-informed batch sampling. On the other hand, PHY-Teacher is a physics-model-based design of action policies that focuses solely on safety-critical functions. PHY-Teacher is novel in its real-time patch for two key missions: i) fostering the teaching-to-learn paradigm for DRL-Student and ii) backing up the safety of real plants. The Trigger manages the interaction between the DRL-Student and the PHY-Teacher. Powered by the three interactive components, the Real-DRL can effectively address safety challenges that arise from the unknown unknowns and the Sim2Real gap. Additionally, Real-DRL notably features i) assured safety, ii) automatic hierarchy learning (i.e., safety-first learning and then high-performance learning), and iii) safety-informed batch sampling to address the learning experience imbalance caused by corner cases. Experiments with a real quadruped robot, a quadruped robot in NVIDIA Isaac Gym, and a cart-pole system, along with comparisons and ablation studies, demonstrate the Real-DRL's effectiveness and unique features.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Safe and constrained RL, one of 15 directions in Deep reinforcement learning methods. Builds on 38 of its 61 references that are in the map. First posted on arXiv 30 Oct 2025.
Builds on
38 of 61 references.
Closest ancestors · 1
Same direction, newest first
Other references · 37
- 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
- 2015Continuous control with deep reinforcement learningcited by 3,407
- 2018Addressing Function Approximation Error in Actor-Critic Methodscited by 2,417
- 2015End-to-End Training of Deep Visuomotor Policiescited by 875
- 2020Deep Reinforcement Learning for Autonomous Driving: A Surveycited by 721
- 2017Sim-to-Real Transfer of Robotic Control with Dynamics Randomizationcited by 672
- 2017Safe Model-based Reinforcement Learning with Stability Guaranteescited by 366
- 2018Sim-to-Real: Learning Agile Locomotion For Quadruped Robotscited by 319
- 2019End-to-End Safe Reinforcement Learning through Barrier Functions for Safety-Critical Continuous Control Taskscited by 289
- 2018Learning to Adapt in Dynamic, Real-World Environments Through Meta-Reinforcement Learningcited by 276
- 2018Residual Reinforcement Learning for Robot Controlcited by 274
- 2018Learning to Drive in a Daycited by 273
25 more
- 2016CAD2RL: Real Single-Image Flight without a Single Real Imagecited by 233
- 2021How to Train Your Robot with Deep Reinforcement Learning; Lessons We've Learnedcited by 213
- 2017Preparing for the Unknown: Learning a Universal Policy with Online System Identificationcited by 141
- 2021Fast and Efficient Locomotion via Learned Gait Transitionscited by 54
- 2021Coupling Vision and Proprioception for Navigation of Legged Robotscited by 51
- 2021Auto-Tuned Sim-to-Real Transfercited by 49
- 2019Control Regularization for Reduced Variance Reinforcement Learningcited by 37
- 2021Stabilizing Neural Control Using Self-Learned Almost Lyapunov Criticscited by 35
- 2021Description of Corner Cases in Automated Driving: Goals and Challengescited by 35
- 2022Safe Reinforcement Learning for Legged Locomotioncited by 35
- 2019Neural Simplex Architecturecited by 29
- 2022Lyapunov Design for Robust and Efficient Robotic Reinforcement Learningcited by 28
- 2019How to pick the domain randomization parameters for sim-to-real transfer of reinforcement learning policies?cited by 25
- 2021Vision-Guided Quadrupedal Locomotion in the Wild with Multi-Modal Delay Randomizationcited by 23
- 2021Bayesian Controller Fusion: Leveraging Control Priors in Deep Reinforcement Learning for Roboticscited by 20
- 2022Learning Minimum-Time Flight in Cluttered Environmentscited by 16
- 2023Posterior Sampling for Deep Reinforcement Learningcited by 16
- 2023Stable and Safe Reinforcement Learning via a Barrier-Lyapunov Actor-Critic Approachcited by 13
- 2023Physics-Regulated Deep Reinforcement Learning: Invariant Embeddingscited by 10
- 2020SL1-Simplex: Safe Velocity Regulation of Self-Driving Vehicles in Dynamic and Unforeseen Environmentscited by 9
- 2021Runtime Safety Assurance for Learning-enabled Control of Autonomous Driving Vehiclescited by 9
- 2020Deep Reinforcement Learning Based Dynamic Route Planning for Minimizing Travel Timecited by 8
- 2022Equipping Black-Box Policies with Model-Based Advice for Stable Nonlinear Controlcited by 6
- 2022Cloud-Edge Training Architecture for Sim-to-Real Deep Reinforcement Learningcited by 5
- 2025Runtime Learning of Quadruped Robots in Wild Environmentscited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2024Safety-Driven Deep Reinforcement Learning Framework for Cobots: A Sim2Real Approach
- 2024Physics-model-guided Worst-case Sampling for Safe Reinforcement Learning
- 2026Explainable Reinforcement Learning via Physics-Aware Policy Distillation
- 2026Safe Reinforcement Learning using Ideas from Model Predictive Control
- 2024Implicit Safe Set Algorithm for Provably Safe Reinforcement Learning
- 2025Assuring the Safety of Reinforcement Learning Components: AMLAS-RL
- 2024xSRL: Safety-Aware Explainable Reinforcement Learning -- Safety as a Product of Explainability
- 2023Towards Safe Autonomous Driving Policies using a Neuro-Symbolic Deep Reinforcement Learning Approach