MARUS - A Marine Robotics Simulator

Lončar, Ivan; Obradović, Juraj; Kraševac, Natko; Mandic, Luka; Kvasić, Igor; Ferreira, Fausto; Slosic, Vladimir; Nađ, Ðula; Mišković, Nikola

doi:10.1109/oceans47191.2022.9976969

Cited by 7 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While there are some publicly available data sets for ship detection [21], [22] with provided experimental results of popular baseline detectors, none of those are applicable to our detection task because of the difference in vessel classes and major visual gap between simulation and real-world images. In addition to that, with no data collection tools in the MBZIRC simulator, we make use of the MARUS simulator [23] and its tools for fast and easy collection of synthetic data sets. In order to use the aforementioned tools, the scene from the MBZIRC simulator had to be recreated in MARUS.…”

Section: Vessel Detectionmentioning

confidence: 99%

A Search Strategy and Vessel Detection in Maritime Environment Using Fixed-Wing UAVs

Peti

Milas

Kraševac

et al. 2023

2023 IEEE Underwater Technology (UT)

View full text Add to dashboard Cite

In this paper, we address the problem of autonomous search and vessel detection in an unknown GNSS-denied maritime environment with fixed-wing UAVs. The main challenge in such environments with limited localization, communication range, and the total number of UAVs and sensors is to implement an appropriate search strategy so that a target vessel can be detected as soon as possible. Thus we present informed and noninformed methods used to search the environment. The informed method relies on an obtained probabilistic map, while the noninformed method navigates the UAVs along predefined paths computed with respect to the environment. The vessel detection method is trained on synthetic data collected in the simulator with data annotation tools. Comparative experiments in simulation have shown that our combination of sensors, search methods and a vessel detection algorithm leads to a successful search for the target vessel in such challenging environments.

show abstract

Section: Vessel Detectionmentioning

confidence: 99%

A Search Strategy and Vessel Detection in Maritime Environment Using Fixed-Wing UAVs

Peti

Milas

Kraševac

et al. 2023

2023 IEEE Underwater Technology (UT)

View full text Add to dashboard Cite

show abstract

“…The CAP-SD system underwent evaluation within a custom Unity [29] underwater simulation environment. Portions of the simulation, such as sensors and communication framework, utilized MARUS [30], while the hydrodynamics components for ASVs and AUVs were implemented using the Dynamic Water Physics [31] asset in Unity. The simu-lation stack, which incorporates a ROS backend, is depicted in Fig 4 . On the ROS [32] side, a gRPC server endpoint is established, functioning as a ROS node responsible for the bidirectional distribution of data flow.…”

Section: A Simulation Architecturementioning

confidence: 99%

Collaborative Damage Detection Framework for Rail Structures Based on a Multi-Agent System Embedded with Soft Multi-Functional Sensors

Cheng

Yao

Yang

et al. 2022

Sensors

View full text Add to dashboard Cite

With the rapid growth of railways in China, the focus has changed to the maintenance of large-scale rail structures. Multi-agent systems (MASs) based on wireless sensor network (WSNs) with soft multi-functional sensors (SMFS) are adopted cooperatively for the structural health monitoring of large-scale rail structures. An MAS framework with three layers, namely the sensing data acquisition layer, sensor data processing layer, and application layer, is built here for collaborative data collection and processing for a rail structure. WSN nodes with strain, temperature, and piezoelectric sensor units are developed for the continuous structural health monitoring of the rail structure. The feature data at different levels are extracted for the online monitoring of the rail structure. Experiments carried out at the Rail Transmit Base at East China Jiaotong University verify that the WSN nodes with SMFS are successfully assembled onto a 100-m-long track for damage detection. Based on the sensing data and feature data, a neural network data fusion agent (DFA) is applied to calculate the damage index value of the track for comprehensive decisions regarding rail damage. The use of WSNs with multi-functional sensors and intelligent algorithms is recommended for cooperative structural health monitoring in railways.

show abstract