In this paper, in order to overcome certain limitations of previously commercialized platforms, a new integrated unmanned surface vehicle (USV) and unmanned underwater vehicle (UUV) platform connected via underwater cable capable of acquiring real-time underwater data and long-time operation are studied. A catamaran-type USV was designed to overcome the limitations of an ocean environment and to play the role as the hub of power supply and communication for the integrated platform. Meanwhile, the UUV was designed as torpedo-shaped to minimize hydrodynamic resistance and its hardware design was focused on processing and sending the underwater camera and sonar data. The underwater cable driven by a winch system was installed to supply power from the USV to the UUV and to transmit acquired data form underwater sonar sensor or camera. Different from other previously studied cooperation system of USVs and autonomous underwater vehicles (AUVs), the merit of the proposed system is real-time motion coordination control between the USV and UUV while transmitting large amount of data using the tether cable. The main focus of the study is coordination of the UUV with respect to the global positioning system (GPS) attached at USV and verification of its performance throughout field tests. Waypoint tracking control algorithm was designed and implemented on USV and relative heading, distance control for USV–UUV coordination was implemented to UUV. To ensure the integrity of the coordination control of the integrated platform, a study on accurate measurement system of the relative position between the USV and the UUV by using the GPS and the ultrashort baseline (USBL) device was performed. Individual tests were conducted to verify the performance of USBL and AHRS, which provide the position and heading data of UUV among the sensors mounted on the actual platform, and the effectiveness of the obtained sensor data is presented. Using the accurate measurement system, a number of field tests were conducted to verify the performance of the integrated platform.

To overcome the limitations of a single unmanned surface vehicle (USV), this study investigated the swarm control algorithms of USVs. Among various swarm control methods, a leader–follower swarm control method was selected and studied. The performance of the swarm algorithm proposed in this study was verified through an actual sea area test. A USV was designed and manufactured by dividing the power and communication parts. The power system was equipped with a coulometer to monitor the battery state in real time to protect the system through a switch linked to the coulometer in the case of abnormalities in the battery and to prevent accidents. In addition, a communication system was established to process the sensor data and camera image data of the USV in real time. Consequently, the desired swarm formation was achieved through the separately constructed swarm control algorithm. Before the actual sea area test, individual performance tests of each sensor were conducted. Finally, the performance of the swarm algorithm was verified by conducting a sea area test. The performance of the controllers was confirmed to be good, and the swarm formation was confirmed to be successful.

This study focuses on faults in the thrusters of unmanned surface vehicles, which are fatal to the integrity of their missions. As for the fault conditions, the breakage of the thruster blade and the entanglement of floating objects were selected, and a data-driven method was used to diagnose the faults. In the data-driven method, it is important to select the sensitive fault feature. In this study, vibration, current consumption, rotational speed and input voltage were selected as fault features. An experiment was conducted in an engineering water tank to obtain and analyze data on fault conditions to verify the validity of the selected features. In addition, a new fault diagnosis algorithm combining principal component analysis and Shannon entropy was applied for analyzing the correlations among fault features. This algorithm reduces the dimensionality of data while preserving their structure and characteristics, and diagnoses faults by quantifying entropy values. A fault is detected by comparing the entropy value and a predetermined threshold value, and is diagnosed by analyzing the entropy value and visualized 2D or 3D principal component results. Moreover, the fault diagnosis performance of the unmanned surface vehicle’s thruster was verified by analyzing the results for each fault condition.

To operate unmanned systems in the ocean, many elements are required. They could be factors related to the equipment performance or operating methods. However, the aforementioned factors are suitable for situations not assuming unexpected events; in the actual ocean, unmanned equipment has the potential to experience problems in operation due to several factors such as environmental or external forces. Having problems in the propulsion system of unmanned marine equipment in unexpected situations affects the execution of duty of the unmanned equipment and connotes the risk of equipment damage and loss. Also, a lot of data generated in the real world is distributed disproportionately for each class, which means that the data of the minor class consists of significantly less data than the data of the majority class, and the normal data, in contrast, anomaly data is a very small amount of data, and there is very little data for learning. Therefore, in this study, anomaly detection based on unsupervised learning was conducted about these problems. In this study, we evaluated anomaly detection based on unsupervised learning to resolve these problems. Based on the results, the state of unmanned systems being operated in the ocean are identified and the anomaly state is examined. In this study, the auto-encoder (AE) method that learns the features of the normal area of the judgment data, that is, a part of machine learning, and the LSTM method for predicting time series states are utilized. This machine learning technology uses time-series learning data on the state of the unmanned system, creates standards to determine normal and anomaly states, and in turn, learns the normal state and determines the anomaly state. Furthermore, the simulation of the proposed method was carried out to verify the algorithm used in this study, and its validity was verified by comparing its data with the measured data. By applying the deduced anomaly detection information to the unmanned surface vehicle (USV) simulator, a fault tolerance control system was designed to enable the execution of duty even in the situation of an anomaly propeller of the USV, and the effectiveness of this study was verified.

Monocular 3D object detection is a challenging task in autonomous systems due to the lack of explicit depth information in single-view images. Existing methods often depend on external depth estimators or expensive sensors, which increase computational complexity and complicate integration into existing systems. To overcome these limitations, we propose AuxDepthNet, an efficient framework for real-time monocular 3D object detection that eliminates the reliance on external depth maps or pre-trained depth models. AuxDepthNet introduces two key components: the Auxiliary Depth Feature (ADF) module, which implicitly learns depth-sensitive features to improve spatial reasoning and computational efficiency, and the Depth Position Mapping (DPM) module, which embeds depth positional information directly into the detection process to enable accurate object localization and 3D bounding box regression. Leveraging the DepthFusion Transformer (DFT) architecture, AuxDepthNet globally integrates visual and depth-sensitive features through depth-guided interactions, ensuring robust and efficient detection. Extensive experiments on the KITTI dataset show that AuxDepthNet achieves state-of-the-art performance, with ${AP}_{3D}$ scores of 24.72% (Easy), 18.63% (Moderate), and 15.31% (Hard), and ${AP}_{BEV}$ scores of 34.11% (Easy), 25.18% (Moderate), and 21.90% (Hard) at an IoU threshold of 0.7.

The purpose of this study is to design a Swarm Control algorithm for the effective mission performance of multiple unmanned surface vehicles (USVs) used for marine research purposes at sea. For this purpose, external force information was utilized for the control of multiple USV swarms using a lead–follow-formation technique. At this time, to efficiently control multiple USVs, the LSTM algorithm was used to learn ocean currents. Then, the predicted ocean currents were used to control USVs, and a study was conducted on behavioral-based control to manage USV formation. In this study, a control system model for several USVs, each equipped with two rear thrusters and a front lateral thruster, was designed. The LSTM algorithm was trained using historical ocean current data to predict the velocity of subsequent ocean currents. These predictions were subsequently utilized as system disturbances to adjust the controller’s thrust. To measure ocean currents at sea as each USV moves, velocity, azimuth, and position data (latitude, longitude) from the GPS units mounted on the USVs were utilized to determine the speed and direction of the hull’s movement. Furthermore, the flow rate was measured using a flow rate sensor on a small USV. The movement and position of the USV were regulated using an Artificial Neural Network-PID (ANN-PID) controller. Subsequently, this study involved a comparative analysis between the results obtained from the designed USV model and those simulated, encompassing the behavioral control rules of the USV swarm and the path traced by the actual USV swarm at sea. The effectiveness of the USV mathematical model and behavior control rules were verified. Through a comparison of the movement paths of the swarm USV with and without the disturbance learning algorithm and the ANN-PID control algorithm applied to the designed simulator, we analyzed the position error and maintenance performance of the swarm formation. Subsequently, we compared the application results.