mlsanityData to 14 Oct 2026

Model-Based Reinforcement Learning Exploits Passive Body Dynamics for High-Performance Biped Robot Locomotion

arXiv 2604.14565 ↗First posted 16 Apr 2026cs.RO cs.SY eess.SYcited by 0

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

Abstract (arXiv)

Embodiment is a significant keyword in recent machine learning fields. This study focused on the passive nature of the body of a biped robot to generate walking and running locomotion using model-based deep reinforcement learning. We constructed two models in a simulator, one with passive elements (e.g., springs) and the other, which is similar to general humanoids, without passive elements. The training of the model with passive elements was highly affected by the attractor of the system. This lead that although the trajectories quickly converged to limit cycles, it took a long time to obtain large rewards. However, thanks to the attractor-driven learning, the acquired locomotion was robust and energy-efficient. The results revealed that robots with passive elements could efficiently acquire high-performance locomotion by utilizing stable limit cycles generated through dynamic interaction between the body and ground. This study demonstrates the importance of implementing passive properties in the body for future embodied AI.

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 13 of its 40 references that are in the map. First posted on arXiv 16 Apr 2026.

Builds on

13 of 40 references.

References · 13

  1. 2020Mastering Atari with Discrete World Modelscited by 1,007
  2. 2018Recurrent World Models Facilitate Policy Evolutioncited by 785
  3. 2019Learning agile and dynamic motor skills for legged robotscited by 504
  4. 2022DayDreamer: World Models for Physical Robot Learningcited by 400
  5. 2021A Survey of Embodied AI: From Simulators to Research Taskscited by 261
  6. 2023Real-World Humanoid Locomotion with Reinforcement Learningcited by 187
  7. 2023ANYmal Parkour: Learning Agile Navigation for Quadrupedal Robotscited by 152
  8. 2023Learning-based Design and Control for Quadrupedal Robots with Parallel-Elastic Actuatorscited by 11
  9. 2022Learning with Muscles: Benefits for Data-Efficiency and Robustness in Anthropomorphic Taskscited by 6
  10. 2023DeepCPG Policies for Robot Locomotioncited by 4
  11. 2023Mechanical Intelligence Simplifies Control in Terrestrial Limbless Locomotioncited by 2
  12. 2022Effect of the Dynamics of a Horizontally Wobbling Mass on Biped Walking Performancecited by 1
1 more
  1. 2023Bipedal Robot Running: Human-like Actuation Timing Using Fast and Slow Adaptationscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024The Duke Humanoid: Design and Control For Energy Efficient Bipedal Locomotion Using Passive Dynamics
  2. 2024Deep Reinforcement Learning for Bipedal Locomotion: A Brief Survey
  3. 2024Reinforcement Learning for Versatile, Dynamic, and Robust Bipedal Locomotion Control
  4. 2024BiRoDiff: Diffusion policies for bipedal robot locomotion on unseen terrains
  5. 2021Adaptive Mimic: Deep Reinforcement Learning of Parameterized Bipedal Walking from Infeasible References ↗
  6. 2024I-CTRL: Imitation to Control Humanoid Robots Through Constrained Reinforcement Learning
  7. 2026Physical Imitation Learning: Distilling Control Policies into Passive Elasticity
  8. 2025RuN: Residual Policy for Natural Humanoid Locomotion