mlsanityData to 14 Oct 2026

E2DRL: End-to-End Differentiable Reinforcement Learning for 3D Obstacle Avoidance in Quadruped Robots

arXiv 2610.19627 ↗First posted 14 Oct 2026cs.ROcited by 0

On the map: Deep reinforcement learning methods → RL for robot navigation and UAV control also fits: RL for robot locomotion and manipulation

Abstract (arXiv)

Differentiable reinforcement learning shows strong promise for robot policy learning, but its application to legged locomotion has remained limited to blind walking or simulated 2D tasks. Extending it to end-to-end visual obstacle avoidance is challenging because hard collision constraints and early termination upon contact corrupt first-order gradients, trapping optimizers in local minima. We present E2DRL, which addresses this with a soft differentiable obstacle-interaction formulation: penetration penalties across multiple body keypoints with velocity-conditioned stop-gradient shaping to provide the first-order training signals for 3D obstacle avoidance. To realize a deployable perception-control system, a privileged teacher trained via Soft Analytic Policy Optimization (SAPO) in a differentiable simulator is distilled into a recurrent depth-based student via DAgger, then blended with a base locomotion policy using a terrain-aware gating module for deployment. In simulation, E2DRL achieves higher sample efficiency than PPO and demonstrates superior velocity-tracking performance and 3D obstacle-avoidance capability compared with both PPO and Duality-based DCBF NMPC. We present the first sim-to-real validation of end-to-end differentiable RL for 3D obstacle avoidance on a Unitree Go2 in static and dynamic environments.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot navigation and UAV control, one of 15 directions in Deep reinforcement learning methods. Builds on 23 of its 30 references that are in the map. First posted on arXiv 14 Oct 2026.

Builds on

23 of 30 references.

References · 23

  1. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,240
  2. 2020Learning Quadrupedal Locomotion over Challenging Terraincited by 604
  3. 2021Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningcited by 512
  4. 2021RMA: Rapid Motor Adaptation for Legged Robotscited by 508
  5. 2019Learning agile and dynamic motor skills for legged robotscited by 504
  6. 2021Brax -- A Differentiable Physics Engine for Large Scale Rigid Body Simulationcited by 407
  7. 2022Learning robust perceptive locomotion for quadrupedal robots in the wildcited by 382
  8. 2023Extreme Parkour with Legged Robotscited by 164
  9. 2023Robot Parkour Learningcited by 149
  10. 2022Walk These Ways: Tuning Robot Control for Generalization with Multiplicity of Behaviorcited by 138
  11. 2022Accelerated Policy Learning with Parallel Differentiable Simulationcited by 104
  12. 2022Do Differentiable Simulators Give Better Policy Gradients?cited by 89
11 more
  1. 2024Agile But Safe: Learning Collision-Free High-Speed Legged Locomotioncited by 73
  2. 2024Back to Newton's Laws: Learning Vision-based Agile Flight via Differentiable Physicscited by 45
  3. 2024Learning Quadruped Locomotion Using Differentiable Simulationcited by 32
  4. 2022Dojo: A Differentiable Physics Engine for Roboticscited by 26
  5. 2024Learning to walk in confined spaces using 3D representationcited by 22
  6. 2024Stabilizing Reinforcement Learning in Differentiable Multiphysics Simulationcited by 20
  7. 2022An Efficient Locally Reactive Controller for Safe Navigation in Visual Teach and Repeat Missionscited by 14
  8. 2024Residual Policy Learning for Perceptive Quadruped Control Using Differentiable Simulationcited by 13
  9. 2022Walking in Narrow Spaces: Safety-critical Locomotion Control for Quadrupedal Robots with Duality-based Optimizationcited by 10
  10. 2024Learning Deployable Locomotion Control via Differentiable Simulationcited by 8
  11. 2026From Distances to Trajectories: Real-Time Signed Distance Function Mapping and Distance-Accelerated Motion Planning for UAVscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Open-DiffLoco: Open-Source Differentiable Learning for Deployable Blind Quadruped Locomotion
  2. 2026Simple but Stable, Fast and Safe: Achieve End-to-end Control by High-Fidelity Differentiable Simulation
  3. 2024Perceptive Pedipulation with Local Obstacle Avoidance
  4. 2026Learning-Based Dynamic Obstacle Avoidance for a UAV Using Only Three Range Sensors
  5. 2026Agile perceptive multi-skill locomotion for quadrupedal robots in the wild
  6. 2025QuadKAN: KAN-Enhanced Quadruped Motion Control via End-to-End Reinforcement Learning
  7. 2025Motion Priors Reimagined: Adapting Flat-Terrain Skills for Complex Quadruped Mobility
  8. 2026Zero-Shot Reactive Obstacle Avoidance for Generative Robot Policies