Cooperative perception among multiple unmanned surface vehicles (USVs) combines complementary observations to extend maritime target sensing beyond the view range and field of a single platform. Developing such systems at scale calls for a unified workflow for configurable multi-USV scenarios, multimodal acquisition, and shared annotations. We present MMUSV-Sim, a perception-oriented maritime simulation and data-generation platform built on Unreal Engine 5 and Project AirSim. It provides island, open-sea, and port environments; configurable weather, time of day, and wave conditions; a diverse vessel asset library; and spline-based multi-vessel motion. MMUSV-Sim acquires RGB, depth, semantic, LiDAR, and radar observations across multiple USVs and captures a common world state for per-agent annotation export. Experiments verify that the configured wave settings produce the intended changes in vessel heave, roll, and pitch, and evaluate the geometric consistency between projected annotations and semantic renderings. In LiDAR-based cooperative BEV vessel detection experiments on the generated multi-USV dataset, Early Fusion achieves an [email protected] of 72.74, compared with 45.54 using a single USV.
Figures & tables
Fig. 1: Motivating cooperative perception scenario in MMUSV-Sim. The target vessel is occluded from USV 1 by the island, while USV 2 and USV 3 observe it from complementary viewpoints. Their observations can be expressed in a shared BEV frame for cooperative perception.
Fig. 2: Overall architecture of MMUSV-Sim. Manual and batch-generated scenario configurations instantiate the UE5 and Project AirSim environment. The control and acquisition interface coordinates vessel motion, multimodal recording while scene evolution is paused, world-state capture, and offline generation of per-agent annotations and structured exports.
Fig. 3: Visual diversity and environmental controllability of MMUSV-Sim. Top row: representative island, port, and open sea topologies, together with the vessel asset library. Middle row: time-of-day variations under a fixed scene configuration. Bottom row: weather variations under a fixed scene configuration.
Fig. 4: Manual scenario configuration and runtime instantiation in MMUSV-Sim. (a) The BEV interface specifies vessel assets, spline trajectories, and environment settings. (b) The corresponding UE5 scene is instantiated from the stored scenario configuration.
Fig. 5: Sensor and annotation visualizations generated by MMUSV-Sim, including surround-view RGB images, depth maps, semantic segmentation, radar PPI, top-view semantic maps, and BEV-projected LiDAR points.
Wave Setting
Heave RMS (m)
Roll RMS ( ∘ )
Pitch RMS ( ∘ )
W0 ( H=0.5 m)
0.0274
0.2054
0.0684
W1 ( H=1.5 m)
0.0818
0.6134
0.2062
W2 ( H=2.5 m)
0.1379
1.0345
0.3395
W3 ( H=3.5 m)
0.1923
1.4421
0.4774
TABLE I: Vessel-pose perturbations under four simulator wave settings.
Scene
Valid Samples
Med. Center Err. (px)
Mask Contain.
Hull Occup.
Island
4,835
2.26
1.000
0.428
Open Sea
1,663
2.07
0.999
0.420
Port
3,856
1.85
0.999
0.420
Overall
10,354
2.08
0.999
0.424
TABLE II: Projected-box coverage and center alignment with top-view semantic renderings.
TABLE III: LiDAR-based cooperative BEV vessel detection on the evaluation subset. AP@ τ uses a BEV IoU threshold of τ ; all metrics are percentages.
Perturbation
Setting
Late Fusion
Early Fusion
Clean
–
55.47
72.74
Delay
200 ms
52.25
69.26
Delay
400 ms
51.33
65.21
Delay
600 ms
46.25
58.06
Pose noise
(0.5m,0.5∘)
48.11
63.64
Pose noise
(1.0m,1.0∘)
40.46
46.94
TABLE IV: Sensitivity of cooperative BEV detection to pose noise and temporal delay. [email protected] values are percentages; pose settings denote (σxyz,σrpy) .
Fig. 6: Qualitative comparison of No Fusion, Late Fusion, and Early Fusion in a port scene. Red and blue boxes denote ground truth and predictions, respectively. In this port scene, Early Fusion better aligns the partially observed large-vessel box.
This paper proposes a cooperative target circumnavigation framework for multiple unmanned surface vehicles (USVs) operating without external localization. The objective is to maintain a uniform circular formation of a specified radius around a target using only limited onboard sensing. The framework adopts a heterogeneous perception strategy that distinguishes between the asymmetric sensing relationships with the target and among the USVs. Specifically, the USVs obtain relative range and displacement measurements through active perception and inter-vehicle communication, while bearing measurements to a non-cooperative target are acquired via passive sensors. To estimate relative positions--both among USVs and between each USV and the target--we employ a Maximum Correntropy Kalman Filter and a Pseudo-Linear Kalman Filter, respectively. A coupled oscillator-based formation controller is designed to ensure system observability while achieving circumnavigation. Theoretical analysis demonstrates that the controller ensures the relative motions between the USVs, as well as that between each USV and the target, satisfy the persistent excitation condition, thereby guaranteeing observability of the Kalman-based filters. The effectiveness of the proposed approach is validated through numerical simulations.
Xueming Liu, Lin Li, Xiang Zhou +2
School of Aeronautics and Astronautics, Sun Yat-sen University (Shenzhen Campus), Shenzhen, China
Offshore maintenance requires operations in the air, the surface, and the subsea domain and include human supervision. This paper presents LOTUSim-Energy, a real-time maritime simulator designed for multi-domain human--drone interaction for offshore operation and maintenance. The plat- form unifies heterogeneous unmanned vehicles (Unmanned Aerial Vehicles: UAVs, Unmanned Surface Vehicles: USVs, Autonomous Underwater Vehicles: AUVs, Remotely Operated Vehicles: ROVs) within a distributed architecture coupling environment forcing (wind, waves, currents) and provides immersive user interfaces for supervision (desktop and virtual reality). A structured offshore task library enables repeatable evaluation of autonomy stacks under realistic metocean disturbances. The simulator supports realistic physics, energy-aware battery modeling, and fault-detection pipelines as modular validation tools. System-level performance is demonstrated on a multi-domain inspection scenario for monopile and transition piece structure, where we evaluate the reliability of integrated waypoint-follower plugin and Automatic Identification System (AIS)-referenced trajectory tracking under real-time energy monitoring. By combining unified environmental physics, heterogeneous vehicle simulation, and immersive supervision, LOTUSim-Energy provides an integration testbed for prototyping and rehearsing offshore human--robot collaboration workflows, as a step toward de-risking sea deployment.
Juliette Grosset, Marie Dubromel, Hélène Lechêne +2
The LiDAR-based multi-agent and single-agent perception has shown promising performance in environmental understanding for robots and automated vehicles. However, there is no existing method that simultaneously solves both multi-agent and single-agent perception in an unsupervised way. By sharing sensor data between multiple agents via communication, this paper discovers two key insights: 1) Improved point cloud density after the data sharing from cooperative views could benefit unsupervised object classification, 2) Cooperative view of multiple agents can be used as unsupervised guidance for the 3D object detection in the single view. Based on these two discovered insights, we propose an Unsupervised Multi-agent and Single-agent (UMS) perception framework that leverages multi-agent cooperation without human annotations to simultaneously solve multi-agent and single-agent perception. UMS combines a learning-based Proposal Purifying Filter to better classify the candidate proposals after multi-agent point cloud density cooperation, followed by a Progressive Proposal Stabilizing module to yield reliable pseudo labels by the easy-to-hard curriculum learning. Furthermore, we design a Cross-View Consensus Learning to use multi-agent cooperative view to guide detection in single-agent view. Experimental results on two public datasets V2V4Real and OPV2V show that our UMS method achieved significantly higher 3D detection performance than the state-of-the-art methods on both multi-agent and single-agent perception tasks in an unsupervised setting.