Learning Natural and Robust Hexapod Locomotion over Complex Terrains via Motion Priors based on Deep Reinforcement Learning
On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation
Abstract (arXiv)
Multi-legged robots offer enhanced stability to navigate complex terrains with their multiple legs interacting with the environment. However, how to effectively coordinate the multiple legs in a larger action exploration space to generate natural and robust movements is a key issue. In this paper, we introduce a motion prior-based approach, successfully applying deep reinforcement learning algorithms to a real hexapod robot. We generate a dataset of optimized motion priors, and train an adversarial discriminator based on the priors to guide the hexapod robot to learn natural gaits. The learned policy is then successfully transferred to a real hexapod robot, and demonstrate natural gait patterns and remarkable robustness without visual information in complex terrains. This is the first time that a reinforcement learning controller has been used to achieve complex terrain walking on a real hexapod robot.
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 18 of its 25 references that are in the map. First posted on arXiv 5 Nov 2025.
Builds on
18 of 25 references.
References · 18
- 2017Proximal Policy Optimization Algorithmscited by 15,215
- 2016Generative Adversarial Imitation Learningcited by 1,597
- 2020Learning Quadrupedal Locomotion over Challenging Terraincited by 604
- 2021Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningcited by 512
- 2021RMA: Rapid Motor Adaptation for Legged Robotscited by 508
- 2019Learning agile and dynamic motor skills for legged robotscited by 504
- 2022Learning robust perceptive locomotion for quadrupedal robots in the wildcited by 382
- 2020Learning Agile Robotic Locomotion Skills by Imitating Animalscited by 251
- 2017Asymmetric Actor Critic for Image-Based Robot Learningcited by 238
- 2018DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skillscited by 163
- 2022Adversarial Motion Priors Make Good Substitutes for Complex Reward Functionscited by 89
- 2022Concurrent Training of a Control Policy and a State Estimator for Dynamic and Robust Legged Locomotioncited by 88
6 more
- 2019Highly Dynamic Quadruped Locomotion via Whole-Body Impulse Control and Model Predictive Controlcited by 54
- 2022Advanced Skills through Multiple Adversarial Motion Priors in Reinforcement Learningcited by 37
- 2016An Efficient Optimal Planning and Control Framework For Quadrupedal Locomotioncited by 12
- 2022BiConMP: A Nonlinear Model Predictive Control Framework for Whole Body Motion Planningcited by 8
- 2020Learning to Walk: Spike Based Reinforcement Learning for Hexapod Robot Central Pattern Generationcited by 6
- 2020Decentralized Deep Reinforcement Learning for a Distributed and Adaptive Locomotion Controller of a Hexapod Robotcited by 4
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2024Experience-Learning Inspired Two-Step Reward Method for Efficient Legged Locomotion Learning Towards Natural and Robust Gaits
- 2023Learning Robust, Agile, Natural Legged Locomotion Skills in the Wild
- 2025Robust Humanoid Walking on Compliant and Uneven Terrain with Deep Reinforcement Learning
- 2025Attention-Based Map Encoding for Learning Generalized Legged Locomotion
- 2025Motion Priors Reimagined: Adapting Flat-Terrain Skills for Complex Quadruped Mobility
- 2025Whole-Body Constrained Learning for Legged Locomotion via Hierarchical Optimization
- 2026Learning Locomotion on Complex Terrain for Quadrupedal Robots with Foot Position Maps and Stability Rewards
- 2025APEX: Action Priors Enable Efficient Exploration for Robust Motion Tracking on Legged Robots ↗