Sim-to-Real Transfer and Robustness Evaluation of Reinforcement Learning Control With Integrated Perception on an ASV for Floating Waste Capture
IEEE Transactions on Field Robotics · 2026 · 3 · 402-424
DOI: 10.1109/tfr.2026.3690128 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Field Robotics, 2026. Verfasst von Luis F. W. Batista und 2 weiteren Autor:innen. Die Arbeit nennt 58 Literaturangaben. Gefördert von: French Agence Nationale de la Recherche through Project Robust Reinforcement Learning for Real-world Autonomous Multitask Agents.
Abstract (Originalsprache)
Autonomous surface vehicles for floating-waste removal operate under varying hydrodynamics, external disturbances, and challenging water-surface perception. We present a field-validated system that combines camera-based polarimetric perception with a lightweight deep-reinforcement-learning (DRL)-based controller for floating-waste detection and capture. Camera detections are converted into water-surface target points and tracked by a controller trained entirely in simulation and deployed directly on a retrofitted autonomous surface vehicle (ASV) platform. Our main contribution is a sim-to-real testing methodology that combines a two-stage simulation protocol with a perception abstraction module designed to mimic real camera behavior, enabling reproducible field trials and explicit evaluation of the sim-to-real gap. We apply this framework in matched simulation and field experiments across 14 disturbance regimes to expose failure modes and evaluate robustness. The results show centimeter-level terminal accuracy and indicate robust control performance under the evaluated perturbation regimes. The main source of degradation is insufficient actuation-model fidelity. We also demonstrate the system in a search-and-capture application using real camera detections in real-world conditions over areas of up to 450 m2. This study distills practical lessons for reliable transfer, including improved actuation-model fidelity, targeted domain randomization (DR), and careful management of latency and timestamps across modules, while highlighting remaining challenges.
Förderung
- French Agence Nationale de la Recherche through Project Robust Reinforcement Learning for Real-world Autonomous Multitask Agents
Zitieren
APA
Batista, L. F. W., Aravecchia, S., & Pradalier, C. (2026). Sim-to-Real Transfer and Robustness Evaluation of Reinforcement Learning Control With Integrated Perception on an ASV for Floating Waste Capture. IEEE Transactions on Field Robotics, 3, 402-424. https://doi.org/10.1109/tfr.2026.3690128
Vancouver
Batista LFW, Aravecchia S, Pradalier C. Sim-to-Real Transfer and Robustness Evaluation of Reinforcement Learning Control With Integrated Perception on an ASV for Floating Waste Capture. IEEE Transactions on Field Robotics. 2026;3:402-424. doi:10.1109/tfr.2026.3690128
BibTeX
@article{Batista2026,
author = {Batista, Luis F. W. and Aravecchia, Stéphanie and Pradalier, Cédric},
title = {Sim-to-Real Transfer and Robustness Evaluation of Reinforcement Learning Control With Integrated Perception on an ASV for Floating Waste Capture},
journal = {IEEE Transactions on Field Robotics},
year = {2026},
volume = {3},
pages = {402--424},
doi = {10.1109/tfr.2026.3690128}
}
EUVIMED