Sim2Swim: Zero-Shot Velocity Control for Agile AUV Maneuvering in 3 Minutes

2025-12-09

AI Summary

Omnidirectional autonomous underwater vehicles (AUVs) face significant challenges in achieving robust, six-degree-of-freedom (6-DOF) agile maneuvering control in complex underwater environments—particularly due to reliance on manual controller tuning and poor generalization across varying payloads and dynamic fluid conditions. Method: This paper proposes a zero-shot simulation-to-reality transfer approach for velocity control based on deep reinforcement learning (DRL), integrating domain randomization with large-scale parallel training exclusively in simulation—requiring no real-world training or post-deployment controller re-tuning. Contribution/Results: The policy converges within three minutes and transfers directly to AUVs of differing configurations. Pool experiments demonstrate robust tracking of time-varying 6-DOF reference trajectories under diverse payloads and environmental disturbances, significantly enhancing agility and adaptability without any fine-tuning.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Intelligent Robots: Behavior Learning & Control

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Holonomic autonomous underwater vehicles (AUVs) have the hardware ability for agile maneuvering in both translational and rotational degrees of freedom (DOFs). However, due to challenges inherent to underwater vehicles, such as complex hydrostatics and hydrodynamics, parametric uncertainties, and frequent changes in dynamics due to payload changes, control is challenging. Performance typically relies on carefully tuned controllers targeting unique platform configurations, and a need for re-tuning for deployment under varying payloads and hydrodynamic conditions. As a consequence, agile maneuvering with simultaneous tracking of time-varying references in both translational and rotational DOFs is rarely utilized in practice. To the best of our knowledge, this paper presents the first general zero-shot sim2real deep reinforcement learning-based (DRL) velocity controller enabling path following and agile 6DOF maneuvering with a training duration of just 3 minutes. Sim2Swim, the proposed approach, inspired by state-of-the-art DRL-based position control, leverages domain randomization and massively parallelized training to converge to field-deployable control policies for AUVs of variable characteristics without post-processing or tuning. Sim2Swim is extensively validated in pool trials for a variety of configurations, showcasing robust control for highly agile motions.

Problem

• Enables agile 6DOF maneuvering for AUVs with zero-shot sim2real control
• Addresses challenges like hydrodynamics and payload changes without retuning
• Trains deployable velocity controllers in just 3 minutes using DRL

Innovation

• Zero-shot sim2real deep reinforcement learning controller
• Domain randomization and massively parallelized training
• Three-minute training for agile 6DOF AUV maneuvering