studies-rov

Studies of ROV concepts 2025 - Part A

4 - Dynamic Target Hunting Under Autonomous Underwater Vehicle (AUV) Motion Planning Based on Improved Dynamic Window Approach (DWA).

Authors: Juan Li, Houtong Lu, Honghan Zhang, and Zihao Zhang A new method for hunting moving targets using multiple Autonomous Underwater Vehicles (AUVs) is proposed. It combines an improved Dynamic Window Approach (DWA) with Rapidly-exploring Random Tree (RRT) and includes collision avoidance rules. These rules help select safe paths, enhancing obstacle avoidance and reliability. A consistency algorithm ensures the task adapts if an AUV fails. The method effectively predicts target movements and allocates hunting points, resulting in efficient multi- AUV hunting strategies.

24 - Hydrodynamics Model Identification and Model-Based Control Application of a New Type of AUV

Authors: Lunyang Lin, Yuxiang Chen, Hong Xiong, Chunliang Yu, Hong Zhu, Yiyang Xing, and Guodong Zheng The hydrodynamic coefficients of underwater robots help assess their maneuverability and stability, aiding in controller design. This paper introduces an Autonomous Underwater Vehicle (AUV) with a negative-lift profile, using a method that integrates neural networks to identify hydrodynamic coefficients. A dynamic model is built for the AUV, and an MPC controller is designed. Simulations and pool tests show that despite errors in calculating hydrodynamic coefficients, the controller still effectively manages the AUV, proving the practicality of using CFD methods for these calculations in real applications.

1 - Coverage path planning for multi-AUV considering ocean currents and sonar performance.

Authors: Xukai Mu and Wei Gao Coverage path planning (CPP) for autonomous unmanned vehicles (AUV) takes into account sonar performance and ocean currents. This study suggests an integrated algorithm that combines the Dijkstra algorithm for the shortest path, Particle Swarm Optimization (PSO), and the ELKAI library to manage multi-AUV task allocation and path optimization, demonstrating effective results in simulations.

29 - Cooperative control for a ROV-based deep-sea mining vehicle with learned uncertain nonlinear dynamics.

Authors: Yuheng Chen, Haicheng Zhang, Weisheng Zou, Haihua Zhang, Daolin Xu This paper proposes an improved deep-sea mining system using a remotely-operated vehicle (ROV) to address track slippage issues in traditional systems. It presents a dynamic model for the ROV-based Deep-sea Mining Vehicle (RDMV) and introduces a cooperative control strategy. A distributed model predictive control (DMPC) controller and a learning-based model predictive control (LMPC) controller are developed, with a prediction function for external disturbances. The effectiveness of the LMPC controller is tested, while the cooperative strategy is validated through simulations.

22 - Design and Performance Verification of A-HFM Signals for Simultaneous Frame Detection, Cell ID Assignment, and Doppler Estimation in AUVs Using Multiple Surface Buoys.

Authors: Sae-Yong Park, Tae-Geon Chung, and Tae-Ho Im This study discusses improvements in Autonomous Underwater Vehicles (AUVs) and their communication capabilities. It introduces a method using Superimposed Adjusted-HFM signals for reliable underwater communication, particularly under Doppler conditions.

23 - ROV Teleoperation in the Presence of Cross-Currents Using Soft Haptics.

Authors: Joshua Brown, Ildar Farkhatdinov, Michael Jenkin The remote operation of underwater vehicles is challenging due to unseen environmental disturbances like cross-currents. This study investigates a new haptic touchpad that uses vibrations to inform operators about cross-currents. An experiment showed that haptic feedback significantly reduced cognitive load and improved positioning accuracy for operators. Although conducted under controlled conditions, this method can be applied in real-world situations to share live environmental data.

5 - A novel real-time intelligent detector for monitoring UAVs in live- line operation on 10 kV distribution networks.

Authors: Haibo Duan, Fanrong Shi, Bo Gao, Yingyue Zhou, Qiushi Cui The live-line operation of 10 kV distribution networks is crucial for steady power supply but faces challenges like poor monitoring and performance. This study introduces the FEM-YOLOv8 algorithm for detecting protective equipment in such operations. Deployed on UAV- compatible edge devices, it enhances monitoring. Key improvements include an upgraded FAST-C2f module for better feature extraction, an efficient channel attention mechanism for detection, and a new loss function for accuracy. The model shows improved precision, recall, and speed.

30 - Model-Based Offline Reinforcement Learning for AUV Path- Following Under Unknown Ocean Currents with Limited Data

Authors: Xinmao Li, Lingbo Geng, Kaizhou Liu, and Yifeng Zhao Minimizing experimental data while ensuring AUV safety in complex underwater environments is crucial. This study proposes a conservative offline model-based Q-learning (CMQL) algorithm for this purpose. CMQL is robust, utilizes data efficiently, and needs no fine-tuning. It uses limited AUV motion data for training, allowing accurate long-term predictions. The system's design ensures strong generalization and disturbance rejection. Simulation results show CMQL effectively follows paths under unknown currents with just 1000 data points, achieving zero-shot transfer for real-world applications.

31 - Target pursuit for multi-AUV system: zero-sum stochastic game with WoLF-PHC assisted.

Authors: Le Hong, Weicheng Cui. Utilizing multiple autonomous underwater vehicles (AUVs) to chase an invading vehicle is complex. This paper presents a system model using a zero-sum stochastic game framework to develop energy-efficient pursuit strategies. A relay-pursuit mechanism and reward function are designed to capture the invading vehicle quickly. The WoLF-PHC algorithm helps AUVs navigate unknown environments, and simulations show the approach's effectiveness and robustness.

25 - Implementation of ROV control from onshore bases: an innovative approach to subsea inspections and interventions.

Authors: Lucas dos Anjos Cristiano Marino, Hernandes Coutinho Fagundes, Vinicius Guedes Gusmão The work proposal aimed to pilot ROV vehicles from onshore control centers to reduce risks and improve decision-making. The team worked with ROV service providers to adapt existing systems and use a private 4G network for control. Challenges arose in network configuration and simulating offshore operations. In early 2023, Fugro and Subsea7 completed several subsea pipeline inspections. The pilot will continue with Oceaneering and explore new transmission technology. The initial phases showed advantages of piloting ROVs from land.

27 - Security of AUV-carried OIRS-assisted quantum key distribution links in underwater channels subject to misalignment.

Authors: Weina Pang, Ping Wang, Bo Bai, Wenwen Chen, Shuang Li, and Kaile Wang To create a secure communication link in underwater networks, quantum key distribution (QKD) protocols are used with underwater wireless optical communication (UWOC) systems. The reliance on line-of-sight creates challenges for this system. This work introduces an optical intelligent reflecting surface (OIRS) on an autonomous underwater vehicle (AUV) to reduce LOS issues and improve reliability for both discrete and continuous- variable QKD.

32 - Multi-Agent Generative Adversarial Interactive Self-Imitation Learning for AUV Formation Control and Obstacle Avoidance

Authors: Zheng Fang, Tianhao Chen, Tian Shen, Dong Jiang, Zheng Zhang, and Guangliang Li Multiple autonomous underwater vehicles (multi-AUVs) can work together to achieve tasks a single AUV cannot. The Multi-Agent Generative Adversarial Imitation Learning (MAGAIL) method has limitations in needing optimal demonstrations. This letter suggests Multi-Agent Generative Adversarial Interactive Self-Imitation Learning (MAGAISIL), which allows AUVs to learn better through self-generated trajectories.

Authors: Yongcheng Cui, Ying Zhang, Cui-Hua Zhang, Simon X. Yang This paper reviews recent research on task cognition and planning for domestic service robots, focusing on technologies like command text parsing, active task cognition, multimodal perception, and action sequence generation. It analyzes challenges in command parsing, explores deep learning enhancements, and discusses research trends and task planning approaches. Finally, it summarizes current challenges and future development directions.

26 - Task cognition and planning for service robots.

6 - Stackelberg game-based anti-disturbance control for unmanned surface vessels via integrative reinforcement learning.

Authors: Yizhen Meng, Chun Liu, Jing Zhao, Jing Huang, Guanbo Jing Unmanned surface vessels (USVs) often face challenges from ocean waves that disrupt their intended path. To address this, a new control strategy using Stackelberg game theory is proposed. It includes a velocity error dynamic system that allows USVs to optimize control and reject disturbances through a leader-follower model. The strategy's effectiveness is confirmed by analyzing Nash equilibrium, and reinforcement learning is used to find the optimal solution. Simulation results show improved tracking accuracy and reduced energy use.

7 - The risk control of rollover safety for new energy vehicles based on wire controlled chassis system

Authors: Guosheng Wang Technological innovation is crucial for the safety of new energy vehicles. A fuzzy dynamic Markov control strategy was used to design an anti-rollover wire-controlled chassis system. The strategy achieved a wheel turning angle error of 0.002° and a body roll angle of 3.0° in sharp turns. The vehicle reached automotive safety integrity level D, reducing accident risk and ensuring passenger safety and dynamic stability. This provides a technical theory for safety control of car rollover.

8 - Maximising the wrench capability of mobile manipulators with experiments on a UVMS

Authors: Wilhelm J. Marais, Oscar Pizarro, and Stefan B. Williams This paper presents methods for optimizing contact wrenches in underwater vehicle manipulator systems (UVMS) using bi-level optimization and linear lower-level problems. It considers multiple contact forces, tracking trajectory, and dynamic motions. Experiments using a 6 degree of freedom underwater robot platform show significant increases in wrench capability compared to existing methods.

9 - Optimization of Configuration Design for Underwater Dam Defect Detection Vehicles

Authors: Hong-Xia Cheng, Mei-Jie Cao, Peng-Fei Xu, Yan Kai, Zi-Peng Wang, Gang Wan, and Qiao Hu This study develops a streamlined underwater vehicle design and operational framework for underwater dam defect detection. The vehicle's design is inspired by bionic principles and uses parametric modeling. Hydrodynamic coefficients are calculated and validated through towing experiments. An optimization function is established with resistance and stability objectives. The vehicle achieves static and dynamic stability in both horizontal and vertical planes, but exhibits limited dynamic stability in the horizontal plane.

10 - Nu—A Marine Life Monitoring and Exploration Submarine System

Authors: Ali A. M. R. Behiry, Tarek Dafar, Ahmed E. M. Hassan, Faisal Hassan, Abdullah AlGohary, and Mounib Khanafer Marine life exploration faces challenges like limited time, depth restrictions, and safety risks. Nu, a 3D-printed Remotely Operated Underwater Vehicle, addresses these by using Long Range technology for wireless operation and a live-feed camera for fish species classification. It uses Brushless Direct Current motors for long-distance movement and water pump motors for precise navigation. Nu's functionality was evaluated in a controlled 2.5-m-deep body of water, achieving an average precision of 60% in fish detection and 97% in species classification.

13 - Securing BGP ASAP: ASPA and other Post-ROV Defenses

Authors: Justin Furuness, Cameron Morris, Reynaldo Morillo, Arvind Kasiliya, Bing Wang, Amir Herzberg The adoption of Route Origin Validation (ROV) is rapidly increasing, making prefix and subprefix hijacks less effective. The Autonomous System Provider Authorization (ASPA) proposal, which focuses on route leakage and foils some other attacks, is being studied. Simulations show that prefix hijacks are less effective than forged-origin hijacks, and subprefix hijacks are reduced. ASPA significantly protects against post-ROV attacks, even in partial adoption, and is effective even when tier-1 ASes are not adopting.

14 - MUFFNet: lightweight dynamic underwater image enhancement network based on multi-scale frequency

Authors: Dechuan Kong, Yandi Zhang, Xiaohu Zhao, Yanqiang Wang, and Lei Cai Underwater Human-Robot Interaction technology has revolutionized marine exploration and conservation. However, basic underwater robots struggle with complex environments, causing blurry images. MUFFNet, an underwater image enhancement network, addresses this issue by utilizing multi-scale frequency analysis. Experimental results show MUFFNet outperforms existing models, consumes fewer computational resources, and aligns enhanced images more closely with human visual perception.

15 - Experimental application of gliders communication system in South China Sea

Authors: Shuyang Jia, Baoheng Liu, Sichen Zou, and Xiaochuan Zhang A water acoustic communication system was designed using underwater gliders. The system consists of a surface wave glider and an acoustic underwater glider, equipped with ultra-short baseline and acoustic modulation devices. Real-time data transmission is performed using underwater acoustic communication equipment. The system can dive to over 1,000m depth and has a standby power consumption of 5 mW.

16 - Adaptive formation learning control for cooperative AUVs under complete uncertainty

Authors: Emadodin Jandaghi, Mingxi Zhou, Paolo Stegagno, and Chengzhi Yuan This paper presents a two-layer framework for adaptive formation control in Autonomous Underwater Vehicles (AUVs), addressing the need for adaptability without knowledge of system dynamics or environmental data. The framework uses a cooperative estimator for inter- agent communication and a decentralized deterministic learning controller for precise trajectory control. The framework's radial basis function neural networks store dynamic information, eliminating the need for relearning after system restarts.

17 - Three-dimensional dynamic trajectory planning for autonomous underwater robots under the PPO-IIFDS Framework

Authors: Joshua Brown, Ildar Farkhatdinov, Michael Jenkin The remote operation of underwater vehicles is challenging due to unseen environmental disturbances like cross-currents. This study investigates a new haptic touchpad that uses vibrations to inform operators about cross-currents. An experiment showed that haptic feedback significantly reduced cognitive load and improved positioning accuracy for operators. Although conducted under controlled conditions, this method can be applied in real-world situations to share live environmental data.

18 - Dynamic Task Allocation for Heterogeneous Multi-Autonomous Underwater Vehicle Collaboration Under Mine Countermeasures Missions

Authors: Juan Li, Baohua Liu, Caiyun Liu, and Cong Lin The SWCBBA-PR algorithm is designed to address the reallocation problem in mine countermeasures (MCM) combat scenarios. It considers underwater communication limitations and introduces a soft time window mechanism and partial reallocation mechanism. The algorithm has been validated through simulation and shows good convergence performance, generating an optimal allocation scheme for MCM mission emergence.

19 - Nonlinear Sliding-Mode Super-Twisting Reaching Law for Unmanned Surface Vessel Formation Control Under Coupling Deception Attacks

Authors: Yifan Wang, Qiang Zhang, Yaping Zhu, Yancai Hu, and Xin Hu This paper presents a nonlinear sliding-mode super- twisting reaching law algorithm designed to address coupling interference in USV formations during cooperative mining operations. The algorithm improves chatter characteristics, addresses low fitting accuracy, and fends off signal-data interference attacks. It also introduces a nonlinear saturation fault-tolerant filtering mechanism and a nonlinear fitting factor. The algorithm's stability is verified through Lyapunov theory, enabling stability within 12s and improving control accuracy for each ship.

20 - Optimized Hydrodynamic Design for Autonomous Underwater Vehicles

Authors: Gang Fan, Xiaojin Liu, Yanan Hao, Guoling Yin, and Long He The study analyzes drag coefficient and lift-to-drag ratio variation in an autonomous underwater vehicle (AUV) using numerical simulation. The response surface method (RSM) and multi-objective genetic algorithm (MOGA) are used to optimize geometric parameters, aiming to improve lift-to-drag ratio and reduce shell mass. The optimization results increase lift-to-drag ratio and shell mass, improving AUV performance and providing a reference for hydrodynamic design.

21 - Autonomous Sea Floor Coverage with Constrained Input Autonomous Underwater Vehicles: Integrated Path Planning and Control

Authors: Athanasios K. Gkesoulis, Panagiotis Georgakis, George C. Karras, and Charalampos P. Bechlioulis This work presents a fully integrated framework for autonomous underwater vehicles (AUVs) to ensure seafloor coverage in adverse environments. It combines optimal coverage path planning with a robust constrained control algorithm. The solution is validated in a high-fidelity simulation environment, demonstrating the synergy between planning and control, making it ready for practical seafloor operations like underwater debris detection.

33 - Design and Simulation of ROV Wheel Electromagnetic Adsorption Structure

Authors: Zihan Sha This paper explores the wheeled magnetic adsorption structure of a remotely operated underwater vehicle (ROV) to replace manual labor in inspecting underwater ferromagnetic surfaces. The design integrates wheeled wall-climbing technology and electromagnets for stable adhesion. The study evaluates electromagnet selection and performance, and simulation experiments validate optimal design parameters.

34 - Design, Development, and Testing of a Compact Underwater ROV for Inspection and Exploration

Authors: Mohd Shahrieel Mohd Aras, Teoh Xi Yin, Fauzal Naim Zohedi, Mohd Bazli Bahar, Mohamad Haniff Harun, Alias Khamis, Zairi Ismael Rizman This study presents the design and development of a four- degree-of-freedom underwater Remotely Operated Vehicle (ROV) for basic underwater inspection tasks. The ROV is designed using Fusion 360 software and evaluated for performance metrics. The final product is a durable, lightweight acrylic frame with sensors, cameras, and weatherproof housing, managed by an Arduino Uno microcontroller.

35 - Research on the Control Method for Remotely Operated Vehicle Active Docking with Autonomous Underwater Vehicles Based on GFSMO-NMPC

Authors: Hongxu Dai, Yunxiu Zhang, Shengguo Cui, Xinhui Zheng, and Qifeng Zhang The study proposes a control method for Remotely Operated Vehicles (ROVs) to actively dock with Autonomous Underwater Vehicles (AUVs), addressing limitations like restricted maneuverability and external disturbances. It designs a process and control strategy, a Nonlinear Model Predictive Controller, and conducts a joint simulation experiment, achieving a 90% success rate under different disturbances.

36 - Ship Resistance Analysis of a Waterjet Remotely Operated Vehicle (ROV) Using Computational Fluid Dynamics (CFD) Method

Authors: Eurico Leonardo Kause, Tri Susanto This study analyzed the resistance of a waterjet remotely operated vehicle (ROV) propulsion system, finding that it increased by 5.69 N using the CFD method and 1.75% in prisoner comparison using the maxsurf method. The waterjet ROV model had a higher thrust increase of 9.21 N compared to a conventional propeller model of 6.94 N. The waterjet ROV's placement on the outer side of the ROV frame released the original propeller, reducing turbulence and friction in the water. However, the surface of the ROV frame with waterjet propulsion components caused higher hydrodynamic pressure.

37 - From Camera Image to Active Target Tracking: Modelling, Encoding and Metrical Analysis for Unmanned Underwater Vehicles

Authors: Samuel Appleby, Giacomo Bergami, and Gary Ushaw Marine mammal monitoring is crucial for cetacean conservation. Traditional tagging methods are intrusive and susceptible to infection. A less intrusive approach uses underwater vehicle navigation (UUV) commanded by a human operator. AI for autonomous underwater vehicle navigation models training environments in simulation, providing visual and physical fidelity suitable for sim-to-real transfer. SWiMM2.0, a Unity simulation modeling of a BLUEROV UUV with a DRL backend, combines simulation modeling with custom behavior metrics for successful tracking.

38 - Deep Reinforcement Learning Based Active Disturbance Rejection Control for ROV Position and Attitude Control

Authors: Gaosheng Luo, Dong Zhang, Wei Feng, Zhe Jiang, and Xingchen Liu This paper proposes a position and attitude control strategy for underwater robots using a reinforcement learning active disturbance rejection controller. The controller has shown satisfactory results in underwater robot control, but fixed-parameter controllers cannot achieve optimal performance. The deep deterministic policy gradient algorithm is used to optimize the linear extended state observer (LESO), allowing for more accurate estimation of model uncertainties and environmental disturbances. Simulation results show the proposed control scheme significantly improves disturbance estimation accuracy, leading to higher control precision and robustness.

39 - Design and mechanical analysis of wall-climbing ROV adsorption structure

Authors: Xuqing Liu, Jie Yu, Xinzhu Li, Changqin Chen, Xiao Chen, Xin Deng This paper explores the design and reliability of a wheeled walking structure using electromagnets for remotely operated underwater vehicles. The structure offers high adsorption reliability, controllable strength, and convenient separation from the magnetic wall. The study also investigates the influence of electromagnet current input on adsorption capacity.

40 - Event-Triggered Adaptive Backstepping Control of Underactuated AUVs with Input Saturation

Authors: Feng Qian, Yusheng Zheng, Ao Wang, and Jianping Cai This paper introduces a dynamic event-triggered adaptive backstepping control for underactuated autonomous underwater vehicle systems (AUVs) with input saturation. The system ensures stability by introducing a new auxiliary signal system, combining the actuated part with the auxiliary signal system, and constructing a dynamic event- triggering mechanism.

41 - Refining features for underwater object detection at the frequency level

Authors: Wenling Wang, Zhibin Yu, and Mengxing Huang Underwater object detection (UOD) is a growing area of research in computer vision. However, existing methods are vulnerable due to light scattering and color shifting. To address these issues, a multi-scale high-frequency information enhancement module and a multi-scale gated channel information optimization module were designed. The method performed better than existing detection methods on three typical underwater object detection datasets, demonstrating its potential in underwater object detection.

42 - Underwater instance segmentation: a method based on channel spatial cross-cooperative attention mechanism and feature prior fusion

Authors: Zhiqian He, Lijie Cao, Xiaoqing Xu, and Jianhao Xu This study proposes a novel Channel Space Coordinates Attention (CSCA) attention module and Channel A Prior Attention Fusion (CAPAF) feature fusion module to improve underwater instance segmentation accuracy in aquaculture. The CSCA module captures local and global information by combining channel and spatial attention weight, while the CAPAF module optimizes feature fusion by removing redundant information. Experimental results show significant improvements in instance segmentation accuracy on various datasets.

43 - Detailed Investigation of Cobalt-Rich Crusts in Complex Seamount Terrains Using the Haima ROV: Integrating Optical Imaging, Sampling, and Acoustic Methods

Authors: Yonghang Li, Huiqiang Yao, Zongheng Chen, Lixing Wang, Haoyi Zhou, Shi Zhang, and Bin Zhao The Haima remotely operated vehicle (ROV) has been used to enhance surveys of cobalt-rich crusts (CRCs) on complex seamount terrains. The 4500-m-class ROV integrates advanced payloads, such as underwater positioning systems, multi-angle cameras, and subsea shallow drilling systems. The data collected from the ROV was highly reliable, with high-definition video mapping of CRC distribution across varied terrains. The Haima ROV successfully investigated CRC resources in complex terrains, laying the groundwork for seamount crust resource evaluations.

44 - Mid-Water Ocean Current Field Estimation Using Radial Basis Functions Based on Multibeam Bathymetric Survey Data for AUV Navigation

Authors: Jiawen Liu, Kaixuan Wang, Shuai Chang, and Lin Pan This paper proposes a method for mid-water ocean current estimation using multibeam bathymetric survey data. It models the regional unknown current field using radius basis functions (RBFs) and establishes an AUV dead-reckoning model. The method also divides multibeam bathymetric point cloud data into submaps and performs a terrain-matching algorithm. Numerical simulations and experiments show that the proposed method accurately captures complex spatial variations in ocean currents, improving positioning accuracy and reducing errors in autonomous underwater vehicle navigation.

45 - Multi-AUV Hunting Strategy Based on Regularized Competitor Model in Deep Reinforcement Learning

Authors: Yancheng Sui, Zhuo Wang, Guiqiang Bai, and Hao Lu This study explores the application of reinforcement learning algorithms in a competitive game scenario involving multi-autonomous underwater vehicles (MUVs). It introduces an optimality operator and redefines the objective function of multi-agent reinforcement learning (MARL), transforming uncertain states into solvable inference problems. The proposed algorithm shows strong adaptability and higher success rate.

46 - The Effective Business Model for Commercialization of ROV Products in Indonesia

Authors: Imam Sutrisno, Budianto , Yuning Widiarti, Mohammad Basuki, Rahmat, Rini Indarti, Dinda Pramanta, Pranowo Sidi This research explores innovative business models for commercializing Remotely Operated Vehicles (ROVs) in the Indonesian market, identifying ROV-as-a-Service, technology licensing, and joint ventures. It highlights the importance of tailoring models to specific market segments and leveraging digital technologies for operational efficiency and customer satisfaction.

47 - Work Systems in Underwater ROV Operations: Educational Insights for Workforce Development in Maritime Engineering

Authors: Minji Yun, Kent J. Crippen This study uses Work Systems Theory to analyze underwater remotely operated vehicle (ROV) operations, identifying ten distinct work systems based on industry sectors, seniority levels, job tasks, and educational requirements. It provides recommendations for engineering education, emphasizing the need for customized programs to meet the evolving workforce demands in this rapidly evolving field.

2 - A hybrid remotely operated underwater vehicle for maintenance operations in aquaculture: Practical insights from Greek fish farms

Authors: Marios Vasileiou, George Vlontzos Aquaculture addresses global seafood demand amid declining wild fish reserves, with underwater vehicles enhancing its expansion by enabling inspections, data collection, and maintenance. This research designs, develops, and evaluates hybrid underwater systems for inspecting and maintaining aquaculture net cages, integrating a manipulator for tasks like debris removal, fish mortality collection, and net repair. A hybrid remotely operated vehicle (ROV/AUV) with software frameworks, mobility, and navigation capabilities is introduced, capable of tetherless autonomous inspection or tethered semi- automatic maintenance. Experimental validations in Greek fish farms confirmed operational success and implications for future applications.

3 - Fault-Tolerant Model Predictive Control for Autonomous Underwater Vehicles Considering Unknown Disturbances

Authors: Yimin Chen, Shaowen Hao, Jian Gao, Jiarun Wang and Le Li This paper introduces a fault-tolerant model predictive control method for autonomous underwater vehicles (AUVs) to maintain heading control despite rudder stuck faults and disturbances. The approach addresses additive faults by incorporating stuck-fault models and decoupling fault terms from control inputs, integrating them into the predictive controller design. Simulation results confirm that the method enables stable navigation and precise heading adjustment in faulty conditions.

11 - TD3 compensated control for uncertain AUV trajectory tracking

Authors: Zhiqun Liu, Chaoyang Chen, Yi Huang, Dan Yang This paper addresses trajectory tracking challenges in autonomous underwater vehicles (AUVs), focusing on issues like underdrive, non-completeness, strong coupling, modeling errors, and random perturbations. A nominal model is first constructed based on AUV dynamics, with control designed using backstepping and Lyapunov methods. To handle modeling inaccuracies and stochastic disturbances, a reinforcement learning compensation method, Twin Delayed Deep Deterministic Policy Gradient (TD3), is introduced. Numerical simulations demonstrate that the proposed control approach significantly mitigates modeling errors and perturbations, improving AUV trajectory tracking accuracy.

12 - In-situ measurement method for large-scale underwater equipment based on dehazing imaging technology

Authors: Shoubo Zhao To address challenges in accurate, long-range in-situ 3D measurement of large-scale underwater equipment, where conventional methods are limited to meter-scale objects, the authors propose a novel underwater dehazing imaging method combining polarization-based dehazing (to suppress water-scattered light via conjugated scattering field modulation) and a multi- frequency sonar-based underwater IGPS (using spherical transmitters for spatial referencing). Leveraging Digital Micromirror Device (DMD) sparse sensing, the integrated system achieves 1 mm accuracy at 100 m range, enabling high-precision deformation monitoring of large submerged structures like dam gates, ships, and pipelines.

28 - Qadralimb: embedded TCP actuators in silicone for grasping tasks and underwater soft robots.

Authors: Vivek Panchal and Yonas Tadesse Recent advances in soft robotics feature Twisted and Coiled Polymer (TCP) actuators embedded in silicone, valued for quiet, low-cost, low-hysteresis, low-voltage operation and customizability. This work characterizes 2- ply TCP actuators in silicone for swimming and grasping, optimizing controllable parameters (voltage, frequency, pre-tension). A four-limb robot (140×60×5 mm³ limbs) achieved 343 mm vertical swim in 640 s. Grasping succeeded in air and water using single or combined limbs. Actuators delivered 12.5% strain at 300 g load and 1.7 kg blocking force with 225 J input. ANSYS FEM simulations matched experiments, establishing novel underwater capability for exploration and hazardous manipulation.