Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.13359 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908801531117568 |
|---|---|
| author | Tunçay, Sümer Andres, Alain Carlucho, Ignacio |
| author_facet | Tunçay, Sümer Andres, Alain Carlucho, Ignacio |
| contents | Autonomous Underwater Vehicles (AUVs) require reliable six-degree-of-freedom (6-DOF) position control to operate effectively in complex and dynamic marine environments. Traditional controllers are effective under nominal conditions but exhibit degraded performance when faced with unmodeled dynamics or environmental disturbances. Reinforcement learning (RL) provides a powerful alternative but training is typically slow and sim-to-real transfer remains challenging. This work introduces a GPU accelerated RL training pipeline built in JAX and MuJoCo-XLA (MJX). By jointly JIT-compiling large-scale parallel physics simulation and learning updates, we achieve training times of under two minutes. Through systematic evaluation of multiple RL algorithms, we show robust 6-DOF trajectory tracking and effective disturbance rejection in real underwater experiments, with policies transferred zero-shot from simulation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_13359 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Fast Policy Learning for 6-DOF Position Control of Underwater Vehicles Tunçay, Sümer Andres, Alain Carlucho, Ignacio Robotics Machine Learning Autonomous Underwater Vehicles (AUVs) require reliable six-degree-of-freedom (6-DOF) position control to operate effectively in complex and dynamic marine environments. Traditional controllers are effective under nominal conditions but exhibit degraded performance when faced with unmodeled dynamics or environmental disturbances. Reinforcement learning (RL) provides a powerful alternative but training is typically slow and sim-to-real transfer remains challenging. This work introduces a GPU accelerated RL training pipeline built in JAX and MuJoCo-XLA (MJX). By jointly JIT-compiling large-scale parallel physics simulation and learning updates, we achieve training times of under two minutes. Through systematic evaluation of multiple RL algorithms, we show robust 6-DOF trajectory tracking and effective disturbance rejection in real underwater experiments, with policies transferred zero-shot from simulation. |
| title | Fast Policy Learning for 6-DOF Position Control of Underwater Vehicles |
| topic | Robotics Machine Learning |
| url | https://arxiv.org/abs/2512.13359 |