mlsanityData to 14 Oct 2026

Dynamic Obstacle Avoidance with Bounded Rationality Adversarial Reinforcement Learning

arXiv 2503.11467 ↗First posted 14 Mar 2025cs.RO cs.LGcited by 0

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

Reinforcement Learning (RL) has proven largely effective in obtaining stable locomotion gaits for legged robots. However, designing control algorithms which can robustly navigate unseen environments with obstacles remains an ongoing problem within quadruped locomotion. To tackle this, it is convenient to solve navigation tasks by means of a hierarchical approach with a low-level locomotion policy and a high-level navigation policy. Crucially, the high-level policy needs to be robust to dynamic obstacles along the path of the agent. In this work, we propose a novel way to endow navigation policies with robustness by a training process that models obstacles as adversarial agents, following the adversarial RL paradigm. Importantly, to improve the reliability of the training process, we bound the rationality of the adversarial agent resorting to quantal response equilibria, and place a curriculum over its rationality. We called this method Hierarchical policies via Quantal response Adversarial Reinforcement Learning (Hi-QARL). We demonstrate the robustness of our method by benchmarking it in unseen randomized mazes with multiple obstacles. To prove its applicability in real scenarios, our method is applied on a Unitree GO1 robot in simulation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 20 of its 23 references that are in the map. First posted on arXiv 14 Mar 2025.

Builds on

20 of 23 references.

References · 20

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
  3. 2021Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningcited by 512
  4. 2017Robust Adversarial Reinforcement Learningcited by 451
  5. 2022Rapid Locomotion via Reinforcement Learningcited by 168
  6. 2023Extreme Parkour with Legged Robotscited by 164
  7. 2022Walk These Ways: Tuning Robot Control for Generalization with Multiplicity of Behaviorcited by 138
  8. 2018Curriculum Adversarial Trainingcited by 89
  9. 2022A Walk in the Park: Learning to Walk in 20 Minutes With Model-Free Reinforcement Learningcited by 83
  10. 2020MushroomRL: Simplifying Reinforcement Learning Researchcited by 50
  11. 2020Self-Paced Deep Reinforcement Learningcited by 50
  12. 2024Learning Robust Autonomous Navigation and Locomotion for Wheeled-Legged Robotscited by 48
8 more
  1. 2023Grow Your Limits: Continuous Improvement with Real-World RL for Robotic Locomotioncited by 24
  2. 2022ViNL: Visual Navigation and Locomotion Over Obstaclescited by 23
  3. 2021A Probabilistic Interpretation of Self-Paced Learning with Applications to Reinforcement Learningcited by 22
  4. 2024Robust Deep Reinforcement Learning Through Adversarial Attacks and Training : A Surveycited by 13
  5. 2023Robust Adversarial Reinforcement Learning via Bounded Rationality Curriculacited by 10
  6. 2022Hierarchical Adaptive Loco-manipulation Control for Quadruped Robotscited by 5
  7. 2023Puppeteer and Marionette: Learning Anticipatory Quadrupedal Locomotion Based on Interactions of a Central Pattern Generator and Supraspinal Drivecited by 5
  8. 2024Learning-based Hierarchical Control: Emulating the Central Nervous System for Bio-Inspired Legged Robot Locomotioncited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Hierarchical Reinforcement Learning and Value Optimization for Challenging Quadruped Locomotion
  2. 2023Learning Robust, Agile, Natural Legged Locomotion Skills in the Wild
  3. 2024Rethinking Robustness Assessment: Adversarial Attacks on Learning-based Quadrupedal Locomotion Controllers
  4. 2024Diffusion-Reinforcement Learning Hierarchical Motion Planning in Multi-agent Adversarial Games
  5. 2024Dexterous Legged Locomotion in Confined 3D Spaces with Reinforcement Learning
  6. 2025Motion Priors Reimagined: Adapting Flat-Terrain Skills for Complex Quadruped Mobility
  7. 2026Learning Highly Dynamic Skills Transition for Quadruped Jumping Through Constrained Space ↗
  8. 2022Advanced Skills through Multiple Adversarial Motion Priors in Reinforcement Learning