MMUSV-Sim: A Perception-Oriented Simulation and Data-Generation Platform for Multi-USV Cooperative Perception

2026-08-14

AI Summary

This study addresses the absence of unified simulation and data generation workflows for multi-USV collaborative perception by proposing a UE5-based multimodal sensing platform. Integrating AirSim with spline-based motion control, the system enables physically accurate wave environment simulation, synchronized multi-sensor acquisition, and automatic global ground truth annotation. Experimental results demonstrate that the platform effectively supports early-fusion collaborative BEV detection research, achieving an AP@0.5 of 72.74, which represents a 27.2 percentage point improvement over single-USV baselines. These findings validate the critical role of unified simulation workflows in facilitating the development and evaluation of collaborative perception algorithms for maritime autonomous systems.

Abstract

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 AP@0.5 of 72.74, compared with 45.54 using a single USV.

Problem

• Multi-USV Cooperative Perception
• Maritime Simulation
• Data Generation
• Multimodal Acquisition
• Shared Annotations

Innovation

• Cooperative Perception
• Multi-USV Simulation
• Data Generation Platform
• Multimodal Sensing
• Shared Annotations