Metraipolis › archive › dc_5d7362e7571f
E2DRL: End-to-End Differentiable Reinforcement Learning for 3D Obstacle Avoidance in Quadruped Robots — arXiv:2610.19627
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.19627
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T21:57:55.309Z
- SHA-256 of the archived text
4a388193c364752c377a68717813e2d702e688b8d1633f35c993ccbe7f9989f4- Stable URL of this page
https://metraipolis.com/documento/dc_5d7362e7571f
How to cite
E2DRL: End-to-End Differentiable Reinforcement Learning for 3D Obstacle Avoidance in Quadruped Robots — arXiv:2610.19627 — arXiv (en), https://arxiv.org/abs/2610.19627 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5d7362e7571f (Metraipolis archive dc_5d7362e7571f, sha256 4a388193c364752c377a68717813e2d702e688b8d1633f35c993ccbe7f9989f4)
BibTeX
@misc{dc5d7362e7571f,
title = {E2DRL: End-to-End Differentiable Reinforcement Learning for 3D Obstacle Avoidance in Quadruped Robots — arXiv:2610.19627},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.19627},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5d7362e7571f; sha256 4a388193c364752c377a68717813e2d702e688b8d1633f35c993ccbe7f9989f4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19627. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
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. Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.RO, nuovi lavori del giorno). Categorie arXiv: Robotics (cs.RO). Pagina del lavoro: https://arxiv.org/abs/2610.19627 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-15T21:38:44.470Z dalla pagina https://arxiv.org/list/cs.RO/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).