ASVSim (AirSim for Surface Vehicles): A High-Fidelity Simulation Framework for Autonomous Surface Vehicle Research
Bavo Lesy, Siemen Herremans, Robin Kerstens, Jan Steckel, Walter Daems, Siegfried Mercelis, Ali Anwar · 2025
Read ASVSim (AirSim for Surface Vehicles): A High-Fidelity Simulation Framework for Autonomous Surface Vehicle Research by Bavo Lesy, Siemen Herremans, Robin Kerstens, Jan Steckel, Walter Daems, Siegfried Mercelis, Ali Anwar (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-22174.txt.
Introduction
Bavo Lesy and six co-authors introduce ASVSim, an open simulator for self-driving boats, in a 2025 paper. Ports and inland waterways in Europe face staff shortages, and greener transport plans push more cargo onto the water.
Good simulators and data sets for training boat systems are scarce. The team adapted an existing air vehicle simulator for surface vessels. Their version models how a boat moves and fakes its radar and camera views. It can also make synthetic data for training vision models and learning agents. Tests on waterway mapping and self-navigation show what the tool can do. The authors offer it as shared ground for both classic control research and deep learning work.
Catalog record
| CREATOR | Bavo Lesy, Siemen Herremans, Robin Kerstens, Jan Steckel, Walter Daems, Siegfried Mercelis, Ali Anwar |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-22174
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
ASVSim (AirSim for Surface Vehicles): A High-Fidelity Simulation Framework for Autonomous Surface Vehicle Research Bavo Lesy, Siemen Herremans, Robin Kerstens, Jan Steckel, Walter Daems, Siegfried Mercelis, Ali Anwar Abstract The transport industry has recently shown significant interest in unmanned surface vehicles (USVs), specifically for port and inland waterway transport. These systems can improve operational efficiency and safety, which is especially relevant in the European Union, where initiatives such as the Green Deal are driving a shift towards increased use of inland waterways. At the same time, a shortage of qualified personnel is accelerating the adoption of autonomous solutions. However, there is a notable lack of open-source, high-fidelity simulation frameworks and datasets for developing and evaluating such solutions. To address these challenges, we introduce AirSim For Surface Vehicles (ASVSim), an open-source simulation framework specifically designed for autonomous shipping research in inland and port environments. The framework combines simulated vessel dynamics with marine sensor simulation capabilities, including radar and camera systems and supports the generation of synthetic datasets for training computer vision models and reinforcement learning agents. Built upon Cosys-AirSim, ASVSim provides a comprehensive platform for developing autonomous navigation algorithms and generating synthetic datasets. The simulator supports research of both traditional control methods and deep learning-based approaches. Through limited experiments, we demonstrate the potential of the simulator in these research areas. ASVSim is provided as an open-source project under the MIT license, making autonomous navigation research accessible to a larger part of the ocean engineering community. [type=editor, orcid=0009-0000-9395-9437 ] [type=editor, orcid=0000-0001-7880-7144] [type=editor, orcid=0000-0002-5127-4947 ] [type=editor, orcid=0000-0003-4489-466X ] [type=editor, orcid=0000-0001-6717-744X ] [type=editor, orcid=0000-0001-9355-6566 ] [type=editor, orcid=0000-0002-5523-0634 ] keywords ,Vessel Simulation,Computer Vision,Unmanned Surface Vehicles ,Open-Source ,Path Planning ,Deep Learning † † credit: Conceptualization, Investigation, Software, Writing - original draft † † credit: Conceptualization, Investigation, Software, Writing - original draft † † credit: Methodology, Software, Writing - original draft † † credit: Funding acquisition, Supervision † † credit: Funding acquisition, Supervision † † credit: Funding acquisition, Supervision † † credit: Writing - review & editing, Funding acquisition, Supervision † † corresponding: Corresponding authors with equal contribution 1 Introduction There is a significant research effort concerning autonomous navigation systems for unmanned surface vehicles (USVs) ( Gu et al. 2021 ; Pietrzykowski et al. 2022 ) . However, developing and testing autonomous navigation systems in real-world environments is challenging due to high costs, safety risks, and the complex nature of maritime operations. For inland waterways, the complexity is further increased by narrow passages, locks, and dense traffic patterns and a frequent need to observe and avoid obstacles through the use of sensors. Infrastructure such as bridges, ports, and shore-based facilities may lead to sensor occlusion, creating blind spots that autonomous systems must account for. Additionally, varying wind and current conditions can significantly impact a vessel, requiring robust control strategies to maintain safe navigation. Because of these challenges, simulation has emerged as a crucial tool for developing and validating navigation algorithms, in a safe and cost-effective manner ( Fossen 2021 ; Zhang et al. 2024 ; Serigstad et al. 2018 ) . For these purposes, it is important that the simulation closely represents the real world to reduce the effort needed to deploy simulated solutions on real vessels. However, a current gap exists in the availability of shipping simulators for autonomy research, with most existing simulators either being designed for human operator training, being proprietary, or lacking the visual realism (or fidelity) needed for accurate camera and sensor simulation. More recently, there is an increasing interest in the potential of deep learning methods in a shipping context, such as obstacle detection ( Jiang et al. 2024 ) or Reinforcement Learning (RL) for navigation ( Herremans et al. 2023 ) . RL has emerged as a promising approach for USVs , partly because of its ability to generalize across different vessel types and operating conditions ( Lesy et al. 2024 ) . However, the USV domain currently lacks accessible, high-fidelity simulation software that can accurately model vessel dynamics and generate realistic sensor data for development of autonomous systems that perceive their environment and navigate within it. The main contribution of this paper is AirSim for Surface Vehicles (ASVSim) 1 1 1 https://github.com/BavoLesy/ASVSim , an open-source, high-fidelity simulator specifically designed for autonomous shipping research. ASVSim is tailored for Inland Waterway Transport (IWT) , where the focus is more on the navigation of vessels through narrow passages and locks, and less on the dynamics of the vessel itself. The simulator currently provides 3-degree-of-freedom (3DoF) maneuvering dynamics, sensor simulations, and environmental conditions such as wind and currents. This enables researchers to develop, train, and test autonomous navigation algorithms in a configurable setting. As will be demonstrated throughout this work, ASVSim allows for the creation of high-fidelity environments, essential for image-based sensor research. Because of this visual fidelity, ASVSim can be used to generate realistic synthetic datasets to train computer vision models. We demonstrate how the simulator can be used for training and validation of autonomy solutions, using examples such as dataset collection and waterway segmentation. Limited autonomous navigation tasks are demonstrated as example use-cases of the simulator, such as autonomous channel navigation and the training of a RL agent to navigate a vessel to a goal whilst avoiding obstacles. The paper is structured as follows: in Section 2 , we review related works in the field of shipping simulators. In Section 3 and 4 , we provide an overview of the mathematical models used in ASVSim , describing vessel dynamics and radar simulation. In Section 5 , we discuss the functionality and practical usage examples of ASVSim , based on the Cosys-AirSim ( Jansen et al. 2023 ) framework, and demonstrate its ease of use through a Python and Matlab application programming interface (API) . Furthermore, in Section 6 , we present the results of our autonomy experiments, including computer vision and path planning tests. Finally, in Section 7 , we provide a conclusion of our work and lay out future research directions. 2 Related Works Table 1: Comparison of Marine Shipping Simulators (excluding underwater simulation) Simulator Open-source High Fidelity Primary Usage SIMTECH ASHSS ( Simtech 2025 ) \times ✓ Autonomous navigation Kongsberg K-Sim ( Kongsberg 2025 ) \times ✓ Maritime training Nautis/Nautis Home ( VSTEP Simulation 2025 ) \times ✓ Maritime training Wärtsilä R&D Simulator ( Wärtsilä 2025 ) \times ✓ Autonomous navigation European Ship Simulator ( Excalibur Games 2016 ) \times ✓ Entertainment uSimMarine (MOOS-IVP) ( Benjamin et al. 2010 ) ✓ \times Autonomous navigation Marine Systems Simulator ( Fossen and Perez 2004 ) ✓ \times Marine control systems ASVSim (ours) ✓ ✓ Autonomous navigation Mission simulation has long been an essential tool in the shipping industry, primarily for training and certification of maritime personnel. The evolution of these simulators has been driven by the need for increasingly realistic training environments that can prepare personnel for operating modern vessels. Industry-standard simulators such as Kongsberg K-Sim ( Kongsberg 2025 ) and Nautis ( VSTEP Simulation 2025 ) have become more and more present in maritime education, providing highly realistic environments with accurate vessel dynamics, environmental conditions, and operational scenarios. These commercial systems excel in their primary use case: training maritime professionals in vessel handling, navigation, and emergency procedures across various vessel