The paper addresses the problem of maritime navigation in coastal areas without access to satellite positioning data. The proposed approach relies on visual information from a monocular camera supported by map-derived coastal information. Although the operational concept assumes the use of publicly available cartographic data, the method’s preparation and evaluation require additional processing steps, including high-resolution aerial imagery, GIS-based extraction of coastal features, semantic segmentation, and trained convolutional neural network models. The problem discussed in the paper concerns, for example, Autonomous Underwater Vehicles that seek to reduce underwater dead-reckoning navigation error by surfacing and using information about what is visible around them, in a way similar to how a human would. To solve the above problem, a system was proposed that compares the camera’s representation of the observed coastline with the map representation of the area where the vehicle is most likely located. The system was validated using real-world data. The tests revealed that the information contained in a flat map is insufficient for accurate position estimation. Accuracy is also significantly affected by errors in the semantic segmentation used to extract land features from camera images, as well as by potential errors in the camera viewing angle. The achieved accuracies are sufficient for navigation away from land, but when operating close to land, the proposed system appears significantly insufficient. The paper specifies the system and reports the results.

This study presents a research methodology and test setup to analyze different types of recreational fins for specific tasks: beginner recreational swimming, deep-water diving, and sports swimming. The fins tested were Aqualung, Scubapro, and Exotech. The primary goal is to present the diving fins research methodology, experimental results, and conclusions. The study achieves high reliability in real-world conditions by combining experimental design with mathematical models based on hydrodynamic similarity. Key to this approach is conducting methodologically sound experiments on a controllable object, allowing precise energy measurements of the swimmer's leg. Despite the challenges posed by the variable thrust force of diving fins in different leg positions, this method remains the most reliable way to integrate experimental data with hydrodynamic theories in comprehensive, multivariate research. Presented results show that fin energy efficiency strongly depends on the interaction between fin geometry, flexural stiffness, and swimming style, with different fin designs achieving maximum efficiency under different operating conditions

This paper addresses the problem of controlling autonomous underwater vehicles (AUVs) operating in a swarm under realistic underwater conditions characterised by inaccurate navigation, limited acoustic communication, and noisy sonar observations. A novel Trail Sonar-Based Algorithm (TSBA) is proposed for leader-follower swarm control. Unlike conventional reactive approaches, TSBA combines sparse acoustic communication with prior knowledge of the mission plan, enabling predictive estimation of the tracked vehicle's state and reducing the dependence on continuous information exchange. To evaluate its effectiveness, TSBA was compared with a machine learning-based controller (NSCSUV) in a simulation environment incorporating navigation drift, sonar measurement errors, and a data-driven model of a real low-cost AUV. The proposed vehicle model achieved a mean speed error of 0.107 m/s and a mean heading error of 14.25°, providing a realistic basis for controller evaluation. Simulation results demonstrated that TSBA consistently outperformed the neural network-based approach in formation keeping while generating smoother control commands and requiring only minimal underwater communication. The algorithm maintained stable swarm behaviour despite sensor inaccuracies and communication constraints. Finally, experiments conducted with a real underwater vehicle confirmed the practical applicability and robustness of the proposed approach under real operating conditions.

This paper presents the design and simulation-based validation of an acoustic navigation system intended for operation within a swarm of underwater vehicles. The system is deployed on a mobile leader unit, while the remaining vehicles, referred to as followers, navigate relative to the leader. The proposed solution utilises two or three acoustic transmitters mounted at the front and rear, and, in the option with three transmitters, also in the middle of the leader platform. These transmitters periodically emit acoustic signals that are received by the follower vehicles. By measuring the time-of-flight of the received signals, followers estimate their distances to the transmitters. This dual(triple)-range information, combined with Kalman filter dead-reckoning, enables relative position estimation with respect to the leader, supporting coordinated swarm movement without reliance on external positioning infrastructure such as GPS, which is unavailable underwater. The system was evaluated in a simulation environment across multiple scenarios with varying levels of distance-measurement error. Rather than modelling detailed acoustic signal propagation, the study focuses on assessing the robustness of the positioning method to measurement inaccuracies. The results demonstrate that the proposed configuration provides useful relative positioning accuracy under a range of error conditions and identifies the operating conditions in which its performance deteriorates, supporting the feasibility of the proposed approach for leader–follower coordination in underwater swarms.

The paper deals with the problem of controlling a swarm of underwater vehicles where vehicle sensors cannot distinguish swarm vehicles from other objects, for example, obstacles. The paper assumes a leader–follower strategy, according to which a swarm of low-cost follower vehicles equipped only with short-range navigation follows the leader along a desired trajectory defined in 2D space. Although the trajectory is defined in 2D space, vehicles can change the depth at which they move, resulting in a temporary loss of sensor data. The paper also considers the possibility of a hybrid swarm consisting of vehicles with different characteristics and a variable formation in which the vehicles move. The vehicles are controlled using neural networks that determine the direction of movement and speed of vehicles and are supplied with data from sensors with different operation ranges. The effectiveness of the networks was verified in simulation conditions. Verification tests have shown that the vehicles are able to: (i) independently form an organized swarm in the initial phase of the operation, (ii) move in two different formations despite interference affecting the operation of sensors and making it difficult to maintain the formation, (iii) cooperate within one swarm despite differences in the maneuverability of individual swarm members, and (iv) change the mission depth as a group; however, it is recommended to carry out the depth change maneuver only on straight sections of the trajectory.

In the paper, a new evolutionary technique called Linear Matrix Genetic Programming (LMGP) is proposed. It is a matrix extension of Linear Genetic Programming and its application is data-driven black-box control-oriented modeling in conditions of limited access to training data. In LMGP, the model is in the form of an evolutionarily-shaped program which is a sequence of matrix operations. Since the program has a hidden state, running it for a sequence of input data has a similar effect to using well-known recurrent neural networks such as Long Short-Term Memory (LSTM) or Gated Recurrent Unit (GRU). To verify the effectiveness of the LMGP, it was compared with different types of neural networks. The task of all the compared techniques was to reproduce the behavior of a nonlinear model of an underwater vehicle. The results of the comparative tests are reported in the paper and they show that the LMGP can quickly find an effective and very simple solution to the given problem. Moreover, a detailed comparison of models, generated by LMGP and LSTM/GRU, revealed that the former are up to four times more accurate than the latter in reproducing vehicle behavior.

The paper presents the application of neural networks evolving under the Hill Climb Assembler Encoding (HCAE) algorithm to control follower autonomous underwater vehicles that are members of a swarm consisting of one leader vehicle and a group of followers. The leader is responsible for global navigation and guiding the swarm, while low-cost followers unable to cover longer distances on their own follow the leader. To locate the leader, the followers only use information about the distance to it. Directional information is unavailable to the followers. Moreover, information about the distance is transmitted to followers with a frequency depending on the number of followers: the more followers, the lower the frequency. In addition to tracking the leader, the followers must also be able to avoid collisions with other followers and the leader. To this end, they are equipped with short-range sensors looking around each follower. The simulations presented in the paper were carried out for different swarm sizes, different sensor ranges, different collision distances, and variable leader speed, and showed high effectiveness of the proposed neural solution. The most common strategy for the followers when following a leader was to circle it at a certain safe distance depending on the sensor range.

Abstract Modularity is a feature of most small, medium and large–scale living organisms that has evolved over many years of evolution. A lot of artificial systems are also modular, however, in this case, the modularity is the most frequently a consequence of a handmade design process. Modular systems that emerge automatically, as a result of a learning process, are very rare. What is more, we do not know mechanisms which result in modularity. The main goal of the paper is to continue the work of other researchers on the origins of modularity, which is a form of optimal organization of matter, and the mechanisms that led to the spontaneous formation of modular living forms in the process of evolution in response to limited resources and environmental variability. The paper focuses on artificial neural networks and proposes a number of mechanisms operating at the genetic level, both those borrowed from the natural world and those designed by hand, the use of which may lead to network modularity and hopefully to an increase in their effectiveness. In addition, the influence of external factors on the shape of the networks, such as the variability of tasks and the conditions in which these tasks are performed, is also analyzed. The analysis is performed using the Hill Climb Assembler Encoding constructive neuro-evolutionary algorithm. The algorithm was extended with various module-oriented mechanisms and tested under various conditions. The aim of the tests was to investigate how individual mechanisms involved in the evolutionary process and factors external to this process affect modularity and efficiency of neural networks.

The paper addresses the problem of modeling complex nonlinear objects such as underwater vehicles. To solve this problem, a new evolutionary technique called Linear Matrix Genetic Programming (LMGP) is proposed which is a matrix extension of Linear Genetic Programming. To model the vehicle, the LMGP uses evolutionary formed program which is a sequence of matrix operations. The program has an input vector, hidden vectors, output vector and a number of matrices. While the program is running, the content of the vectors is modified by various operations on them, e.g. addition, multiplication, the change according to sigmoid or hyperbolic tangent functions. Since the program has a hidden state, running it for a sequence of input data has a similar effect to using a recurrent neural network such as LSTM or GRU. To evolve programs, Hill Climb Modular Assembler Encoding is applied. The evolution decides about the number and type of operations in the program, parameters of the operations, and the content of matrices. The program itself is also encoded in the form of a matrix. In order to verify the effectiveness of the proposed techique it was compared with different types of neural networks. The task of all the compared techniques was to reproduce the behavior of a nonlinear model of underwater vehicle. The results of the comparative tests are reported in the paper and they show that the proposed algorithm is able to quickly find an effective and very simple solution to the given problem.

The paper presents the application of an evolutionary recurrent neural network to control the swarm of underwater vehicles. In the swarm, one vehicle is the leader and the others are followers. The leader leads the swarm along a predefined trajectory without regard for the followers while the followers follow the leader and avoid collisions with all other vehicles. Avoiding collisions by the swarm with external obstacles is done by changing the depth. The leader is responsible for detecting the obstacles and informing all the followers about the need to change the depth. To follow the leader, the followers use the information about the distance to it. Directional information is unavailable to them. To avoid collisions inside the swarm, the followers use shortrange sensors.

The paper presents a neural control system for a swarm of underwater vehicles. A swarm consists of a leader vehicle and follower vehicles. The leader leads the swarm along a predetermined trajectory, detects obstacles and determines obstacle avoidance manoeuvrers for the whole swarm, while the followers follow the leader in a specific formation and avoid the leader and each other. In order to keep the formation without colliding with other members of the swarm, the followers are equipped with a neural control system designed with a neuro-evolutionary algorithm called Hill Climb Assembler Encoding. The system is fed with three sources of information. The first is the leader which, using an acoustic communication system, sends information about its manoeuvrers as well as about the desired type of formation. The second source is sonar, i.e. underwater acoustic radar, whose disadvantage is the difficulty in interpreting the results of operation, i.e. it is difficult to determine whether the observed object is another vehicle or an obstacle. The third source is a camera observation unit consisting of three cameras located around the vehicle. Unfortunately, the disadvantage of this unit is a very short range in opaque water. The proposed control system was tested in simulation conditions and proved to be highly effective. The results of the simulation as well as the design of the system are presented in the paper.

In order for any construction to arise on the sea bottom, the area around has to be carefully cleaned from all dangerous objects including ferromagnetic objects like ammunition or the remains of sea mines. Magnetic anomaly detectors (MAD) which are composed of a set of magnetometers are applied to detect such objects. The combined signal of all the magnetometers moving near the sea bottom is a base for detecting anomalies in the Earth magnetic field produced by ferromagnetic objects. To carry the magnetometers, underwater vehicles can be used. However, in order for the vehicles to be able to fulfill the role of MAD, they have to move near the sea bottom in a tight swarm formation. In the paper, a modular neuro-evolutionary controller is presented which leads each swarm member along a desired trajectory, with the time synchronization, and at the certain distance from the sea bottom. In order to design the controller, two neuro-evolutionary algorithms were applied, i.e. Hill Climb Modular Assembler Encoding (HCMAE) and Differential Evolution (DE). During tests carried out in simulation conditions, initially in 2D and then also in a 3D environment, under the influence of sea current, neuro-controllers evolved by HCMAE and DE were compared with algorithmic and PID-like controllers. The tests showed that the solutions proposed in the paper are more effective than the reference solutions and proved that even in unfavorable underwater conditions it is possible to precisely keep the trajectory specified both spatially and temporally.

The paper presents a neuro-evolutionary algorithm called Hill Climb Assembler Encoding (HCAE) which is a light variant of Hill Climb Modular Assembler Encoding (HCMAE). While HCMAE, as the name implies, is dedicated to modular neural networks, the target application of HCAE is to evolve small/mid-scale monolithic neural networks which, in spite of the great success of deep architectures, are still in use, for example, in robotic systems. The paper analyses the influence of different mechanisms incorporated into HCAE on the effectiveness of evolved neural networks and compares it with a number of rival algorithms. In order to verify the ability of HCAE to evolve effective small/mid-scale neural networks, both feed forward and recurrent, it was tested on fourteen identification problems including the two-spiral problem, which is a well-known binary classification benchmark, and on two control problems, i.e., the inverted-pendulum problem, which is a classical control benchmark, and the trajectory-following problem, which is a real problem in underwater robotics. Four other neuro-evolutionary algorithms, four particle swarm optimization methods, differential evolution, and a well-known back-propagation algorithm, were applied as a point of reference for HCAE. The experiments reported in the paper revealed that the evolutionary approach applied in the proposed algorithm makes it a more effective tool for solving the test problems than all the rivals.

The paper presents a novel generative Neuro-Evolutionary (NE) method called Hill Climb Modular Assembler Encoding (HCMAE). The target application of the HCMAE is to evolve modular Artificial Neural Networks (ANNs) whose modular structure is known in advance. Different variants of HCMAE were tested on two well-known ANN benchmarks, i.e. the Two-Spiral problem (feed-forward ANNs), and the Inverted-Pendulum problem (recurrent ANNs), for four different modular neural architectures. Particle Swarm Optimization and Differential Evolution were selected as rivals for HCMAE. Both rival methods were tested in two variants, i.e. a classical one-population variant and cooperative co-evolutionary multi-population variant. The paper presents the proposed method and reports all the experiments. (C) 2021 Elsevier B.V. All rights reserved.

Autonomous underwater vehicles (AUVs) are underwater robots which are able to perform certain tasks without the help of a human operator. The key skill of each AUV is the capability to avoid collisions. To this end, appropriate devices and software are necessary with the potential to detect obstacles and to take proper decisions from the point of view of both the task and safety of the vehicle. The paper presents a neural collision avoidance system (NCAS) designed for the biomimetic autonomous underwater vehicle (BAUV). The NCAS is a component of the path following and collision avoidance system (PFCAS), which as the name implies is responsible for safely leading the vehicle along a desired path with collision avoidance. The task of NCAS is to make decisions regarding vehicle maneuvers in the horizontal plane, but only in the close proximity of the obstacles. It is implemented as an evolutionary artificial neural network designed by means of a neuro-evolutionary technique called assembler encoding with evolvable operations (AEEO). The paper outlines operation and construction of the BAUV as well as the PFCAS, the role of the NCAS in the entire system, and briefly presents AEEO as well as reporting on the experiments performed in simulation.

Autonomous underwater vehicles (AUVs) are robots that operate in underwater environment and do not need involvement of an operator when performing some tasks. In order to move independently in water environment, AUVs need navigation capabilities, on the one hand, they have to be able to detect obstacles and avoid them, and on the other hand, they also have to know their own position and spatial orientation, at least course. With regard to the orientation, there are many various solutions like inertial systems, inclinometers, magnetic compasses, optical gyro–compasses, whereas, position due to unavailability of GPS requires solutions dedicated to underwater environment such as inertial navigation. To this end, information about spatial orientation and velocity is necessary. When the vehicle is not equipped with a device to measure velocity, e.g. because of small size of the vehicle itself, the only solution is to use odometry, that is, to apply information from the drive to estimate the velocity. The paper presents Odometric Navigational System (ONS) designed for a small biomimetic autonomous underwater vehicle (BAUV) and tuned by means of neuro–evolutionary techniques. To verify system performance, data from the real BAUV were applied.

This paper compares a number of state-of-the-art anomaly detection methods on real ship trajectories obtained by an Automatic Identification System (AIS) in the Baltic sea. Because most methods need fixed length trajectory representations, the paper also gives some solutions for reducing variable length trajectories to a fixed size.

The paper presents a concept of the optical coastal navigational system. The task of the system is to provide information about object position in coastal areas based on optical information. The system is intended for Autonomous Underwater Vehicles that operate in GPS denied environments. In addition to the concept itself, the paper also outlines the first results achieved during the research on the system.

The information about ship spatial orientation is required by different ship systems. For example, dynamic positioning, artillery, or video tracking systems, to stabilize a ship, a gun or a camera, have to work in advance, and therefore, they need the information about future ship orientation. To obtain this information, a current ship orientation and sometimes also historical orientations are necessary. Another example is ship echo-sounders applied to hydro-graphic surveys. Since each list of the ship affects echo-sounder depth measurement, to determine true depth of the sea based on the echo-sounders, the spatial orientation of the ship at the moment of each measurement has to be known. One method for determining the ship spatial orientation is an optical system including video cameras, a computer and specialized software. The system works through extracting the horizon line from each video shot and calculating an angle between the line and the horizontal border of the shot. The main disadvantage of this approach is its slowness. There are a number of methods which can be used to extract the horizon line from an image; however, they are so computationally complex that their application to high-resolution images is limited rather to only post-processing. In the paper, a quick method for horizon line detection is proposed. The experiments on real marine images showed that it is at least as accurate as other methods, what is more, it is also a number of times faster and it seems to be more reliable than its rivals.