Este trabalho trata do estudo de concepções de arquitetura do controle aplicadas aos robôs móveis autônomos e da proposição de um delas à instrumentação e controle em tempo real de um modelo de embarcação naval de alto desempenho. Tal veículo remotamente operado foi desenvolvido como parte das atividades do projeto temático "Comportamento em Ondas de Embarcações de Alto Desempenho" (proc.Fapesp 1997/13090-3). Realizou-se uma investigação dos diversos paradigmas de inteligência artificial que orientaram a evolução dos robôs móveis autônomos até o presente momento e, em particular, as concepções baseadas em modelos sócio-antropológicos e computacionais (teoria de agentes e orientação a objetos) através de sua aplicação à implementação de um sistema de aquisição e controle orientado a objetos, modelado através da UML (Unified Modeling Language), para o veículo mencionado. Testes de validação da arquitetura do controle foram realizados, sendo obtidos resultados experimentais que permitiram análises a respeito da dinâmica, manobrabilidade e navegação do veículo, as quais sugerem vários aperfeiçoamentos para o sistema de hardware e software em trabalhos futuros. / This work deals with the study of control architecture approaches applied to autonomous mobile robots, and proposes one of them for the control system of a self-propelled high speed ship model. Such unmanned vehicle was developed for the research project “Comportamento em Ondas de Embarcações de Alto Desempenho” (proc. FAPESP 1997/13090-3). A number of artificial intelligence paradigms, related to the autonomous robot evolution up to now, were investigated. Models based on the socio-anthropological paradigm and the corresponding computer science approaches, i.e. agent theory and object-oriented modeling, were emphasized. Object-oriented control software based on the UML (Unified Modeling Language) was designed for the real-time embedded system of the ship model. Validation tests of the control architecture were carried out. Experimental results, related to vehicle dynamics, maneuverability and navigation were acquired by the embedded system and analyzed in this work. These results suggest a number of improvements for future works on the software and hardware systems.

Este trabalho trata do estudo de concepções de arquitetura do controle aplicadas aos robôs móveis autônomos e da proposição de um delas à instrumentação e controle em tempo real de um modelo de embarcação naval de alto desempenho. Tal veículo remotamente operado foi desenvolvido como parte das atividades do projeto temático "Comportamento em Ondas de Embarcações de Alto Desempenho" (proc.Fapesp 1997/13090-3). Realizou-se uma investigação dos diversos paradigmas de inteligência artificial que orientaram a evolução dos robôs móveis autônomos até o presente momento e, em particular, as concepções baseadas em modelos sócio-antropológicos e computacionais (teoria de agentes e orientação a objetos) através de sua aplicação à implementação de um sistema de aquisição e controle orientado a objetos, modelado através da UML (Unified Modeling Language), para o veículo mencionado. Testes de validação da arquitetura do controle foram realizados, sendo obtidos resultados experimentais que permitiram análises a respeito da dinâmica, manobrabilidade e navegação do veículo, as quais sugerem vários aperfeiçoamentos para o sistema de hardware e software em trabalhos futuros. / This work deals with the study of control architecture approaches applied to autonomous mobile robots, and proposes one of them for the control system of a self-propelled high speed ship model. Such unmanned vehicle was developed for the research project “Comportamento em Ondas de Embarcações de Alto Desempenho" (proc. FAPESP 1997/13090-3). A number of artificial intelligence paradigms, related to the autonomous robot evolution up to now, were investigated. Models based on the socio-anthropological paradigm and the corresponding computer science approaches, i.e. agent theory and object-oriented modeling, were emphasized. Object-oriented control software based on the UML (Unified Modeling Language) was designed for the real-time embedded system of the ship model. Validation tests of the control architecture were carried out. Experimental results, related to vehicle dynamics, maneuverability and navigation were acquired by the embedded system and analyzed in this work. These results suggest a number of improvements for future works on the software and hardware systems.

Cette thèse vise à rendre possible l’utilisation des véhicules sous-marins autonomes (AUVs) dans l’inspection visuelle des structures offshore. Deux tâches sont identifiées: le suivi des structures rectilignes et la stabilisation devant les cibles planaires. Les AUVs complétement actionnés et équipés d'une centrale inertielle, un DVL, un altimètre et une caméra vidéo sont visés. La dynamique en 6 d.d.l. d'un AUV est rappelée. L'architecture de contrôle reflétant la structure en cascade de la dynamique est choisie. Une boucle interne asservie la vitesse du véhicule à la consigne et une boucle externe calcule la vitesse de référence à partir des informations visuelles. Le suivi de pipe est assuré par l'asservissement visuel 2D qui calcule la vitesse de référence à partir des bords du pipeline détectés dans l’image. La convergence globale asymptotique et localement exponentielle de la position, de l’orientation et de la vitesse sont obtenues. Le contrôleur de stabilisation utilise la matrice d’homographie. Seule la connaissance imprécise de l’orientation de la cible est nécessaire. L’information cartésienne de la profondeur de la cible est estimée à l’aide d’un observateur. La convergence quasi-globale et localement exponentielle de ce contrôleur est démontrée. Afin de tester ces méthodes un simulateur a été développé. Des images de synthèse de haute-fidélité sont générées à partir de simulateur Morse. Elles sont traitées en temps réel à l’aide de la bibliothèque OpenCV. Un modèle Simulink calcule la dynamique complète des 6 d.d.l. du véhicule simulé. Des résultats détaillés sont présentés et mettent en avant les résultats théoriques obtenus. / This thesis deals with a control system for a underwater autonomous vehicle given a two consequent tasks: following a linear object and stabilisation with respect to a planar target using an on-board camera. The proposed solution of this control problem takes advantage of a cascading nature of the system and divides it into a velocity pilot control and two visual servoing schemes. The serving controllers generate the reference velocity on the basis of visual information; line following is based on binormalized Pluecker coordinates of parallel lines corresponding to the pipe contours detected in the image, while the stabilisation relies on the planar homography matrix of observed object features, w.r.t. the image of the same object observed at the desired pose. The pilot, constructed on the full 6 d.o.f. nonlinear model of the AUV, assures that the vehicle’s linear and angular velocities converge to their respective setpoints. Both image servoing schemes are based on minimal assumptions and knowledge of the environment. Validation is provided by a high-fidelity 6 d.o.f. dynamics simulation coupled with a challenging 3D visual environment, which generates images for automatic processing and visual servoing. A custom simulator is built that consist of a Simulink model for dynamics simulation and the MORSE robot and sensor simulator, bound together by ROS message passing libraries. The OpenCV library is used for real-time image processing. Methods of visual data filtering are described. Thus generated experimental data is provided that confirms the desired properties of the control scheme presented earlier.

Today more and more different types of autonomous robots and vehicles are being developed. Most of these rely on the global positioning system and/or communication with other robots and vehicles to determine their global position. However, these are not viable options for the autonomous underwater vehicles (AUVs) of today since radio-waves does not travel well in water. Instead, various techniques for determining the AUVs position are used which comes with a margin of error. This thesis examines the problem of efficiently performing a local search within this margin of error with the objective of finding a docking-station or a bouy.To solve this problem research was made on the subject of search theory and how it previously has been applied in this context. What was found was that classical bayesian search theory had not been used very often in this context since it would require to much processing power to be a viable option in the embedded systems that is AUVs. Instead different heuristics were used to get solutions that still were viable for the situations in which they were used, even though they maybe wasn’t optimal.Based on this the search-strategies Spiral, Greedy, Look-ahead and Quadtree were developed and evaluated in a simulator. Their mean time to detection (MTTD) were compared as well as the average time it took for the strategies to process a search. Look-ahead was the best one of the four different strategies with respect to the MTTD and based on this it is suggested that it should be implemented and evaluated in a real AUV. / Idag utvecklas allt fler olika typer av autonoma robotar och fordon. De flesta av dessa är beroende av det globala positioneringssystemet och/eller kommunikation med andra robotar och fordon för att bestämma deras globala position. Detta är dock inte realistiska alternativ för autonoma undervattensfordon (AUV) idag eftersom radiovågor inte färdas bra i vatten. I stället används olika tekniker för att bestämma AUVens position, tekniker som ofta har en felmarginal. Denna rapport undersöker problemet med att effektivt utföra en lokal sökning inom denna felmarginal med målet att hitta en dockningsstation eller en boj.För att lösa detta problem gjordes en litteraturstudie om ämnet sökteori och hur det tidigare har tillämpats i detta sammanhang. Det som hittades var att den klassiska bayesiska sökteorin inte hade använts mycket ofta i detta sammanhang eftersom det skulle kräva för mycket processorkraft för att det skulle vara ett rimligt alternativ för de inbyggda systemen på en AUV. Istället användes olika heuristiska metoder för att få lösningar som fortfarande var dugliga för de situationer där de användes, även om de kanske inte var optimala.Baserat på detta utvecklades sökstrategierna Spiral, Greedy, Look-ahead och Quad-tree och utvärderades i en simulator. Deras genomsnittliga tid för att upptäcka målet (MTTD) jämfördes liksom den genomsnittliga tiden det tog för strategierna att bearbeta en sökning. Look-ahead var den bästa av de fyra olika strategierna med avseende på MTTD och baserat på detta föreslås det att den ska implementeras och utvärderas i en verklig AUV.

Autonomous underwater vehicle (AUV) navigation has been a challenging problem for a long time. Navigation is challenging due to the drift present in underwater environments and the lack of precise localisation systems such as GPS. Therefore, the uncertainty of the vehicle’s pose grows with the mission’s duration. This research investigates methods to form constraints on the vehicle’s pose throughout typical surveys. Current underwater navigation relies on acoustic sensors. Side Scan Sonar (SSS) is cheaper than Multibeam echosounder (MBES) but can generate 2D intensity images of wide sections of the seafloor instead of 3D representations. The methodology consists in extracting information from pairs of side-scan sonar images representing overlapping portions of the seafloor and computing the sensor pose transformation between the two reference frames of the image to generate constraints on the pose. The chosen approach relies on optimisation methods within a Simultaneous Localisation and Mapping (SLAM) framework to directly correct the trajectory and provide the best estimate of the AUV pose. I tested the optimisation system on simulated data to evaluate the proof of concept. Lastly, as an experiment trial, I tested the implementation on an annotated dataset containing overlapping side-scan sonar images provided by SMaRC. The simulated results indicate that AUV pose error can be reduced by optimisation, even with various noise levels in the measurements. / Navigering av autonoma undervattensfordon (AUV) har länge varit ett utmanande problem. Navigering är en utmaning på grund av den drift som förekommer i undervattensmiljöer och bristen på exakta lokaliseringssystem som GPS. Därför ökar osäkerheten i fråga om fordonets position med uppdragets längd. I denna forskning undersöks metoder för att skapa begränsningar för fordonets position under typiska undersökningar. Nuvarande undervattensnavigering bygger på akustiska sensorer. Side Scan Sonar (SSS) är billigare än Multibeam echosounder (MBES) men kan generera 2D-intensitetsbilder av stora delar av havsbotten i stället för 3D-bilder. Metoden består i att extrahera information från par av side-scan sonarbilder som representerar överlappande delar av havsbotten och beräkna sensorns posetransformation mellan bildens två referensramar för att generera begränsningar för poset. Det valda tillvägagångssättet bygger på optimeringsmetoder inom en SLAM-ram (Simultaneous Localisation and Mapping) för att direkt korrigera banan och ge den bästa uppskattningen av AUV:s position. Jag testade optimeringssystemet på simulerade data för att utvärdera konceptet. Slutligen testade jag genomförandet på ett annoterat dataset med överlappande side-scan sonarbilder från SMaRC. De simulerade resultaten visar att AUV:s poseringsfel kan minskas genom optimering, även med olika brusnivåer i mätningarna.

[pt] No campo em crescimento da robótica subaquática, Veículos Subaquáticos Automatizados (AUVs) estão se tornando cada vez mais importantes para uma variedade de usos, como exploração, mapeamento e inspeção. Esta dissertação foca em estudar os principais desafios da acoplagem automática de AUVs, considerando um ambiente 3D simulado personalizado. A pesquisa divide essa tarefa em duas partes principais: estimativa da pose da garagem e estratégia de controle do AUV. Utilizando uma mistura de métodos tradicionais e novos, incluindo sistemas baseados em marcos fiduciais, Redes Neurais Convolucionais (CNN) e Aprendizado por Reforço (RL), o estudo realiza experimentos para verificar o desempenho e as limitações do sistema. Um aspecto significativo desta dissertação é o uso de um ambiente 3D simulado para facilitar o desenvolvimento e o teste de algoritmos de acoplagem automática para AUVs. Este ambiente simula dinâmicas subaquáticas, sensores robóticos e atuadores, permitindo experimentar diferentes técnicas de estimativa de pose e estratégias de controle. Além disso, o estabelecimento de um ambiente 3D simulado amigável para RL representa uma contribuição relevante, oferecendo uma plataforma reutilizável que não apenas valida os algoritmos de acoplagem automática desenvolvidos neste estudo, mas também serve como base para futuras aplicações subaquáticas baseadas em RL. Em resumo, a dissertação explora uma série de cenários para avaliar a eficácia de várias técnicas de acoplagem automática. Inicialmente, ela utiliza servo-visualização junto com um controlador PID tradicional, seguido pela introdução de métodos mais avançados, como estimadores de pose baseados em CNN e controladores de Aprendizado por Reforço. Esses métodos são avaliados tanto individualmente quanto em combinações híbridas para medir sua adequação e limitações para entender os principais desafios por trás da acoplagem automática de AUVs. / [en] In the growing field of underwater robotics, Automated Underwater Vehicles (AUVs) are becoming more important for a range of uses, such as exploration, mapping, and inspection. This dissertation focuses on studying the main challenges of AUV auto-docking, considering a customized 3D simulated environment. The research breaks down this challenging task into two main parts: cage pose estimation and AUV control strategy. Using a mix of traditional and new methods, including fiducial-based systems, Convolutional Neural Networks (CNN), and Reinforcement Learning (RL), the study carries out experiments to check system performance and limitations. A significant aspect of this dissertation is using a 3D simulated environment to facilitate the development and testing of auto-docking algorithms for AUVs. This environment simulates crucial underwater dynamics, robotic sensors, and actuators, allowing for experimenting with different pose estimation techniques and control strategies. Additionally, the establishment of an RL-friendly 3D simulated environment stands as a relevant contribution, offering a reusable platform that not only validates the auto-docking algorithms developed in this study but also serves as a foundation for future RL-based underwater applications. In summary, the dissertation explores a range of scenarios to evaluate the efficacy of various auto-docking techniques. It initially utilizes visual servoing along with a traditional PID controller, followed by the introduction of more advanced methods like CNN-based pose estimators and Reinforcement Learning controllers. These methods are assessed both individually and in hybrid combinations to gauge their suitability and limitations for understanding the main challenges behind the AUV auto-docking.

In this thesis, we explore Lagrangian Neural Networks (LNNs) for system identification of Autonomous Underwater Vehicles (AUVs) with 6 degrees of freedom. One of the main challenges of AUVs is that they have limited wireless communication and navigation under water. AUVs operate under strict and uncertain conditions, where they need to be able to navigate and perform tasks in unknown ocean environments with limited and noisy sensor data. A crucial requirement for localization and adaptive control of AUVs is having an accurate and reliable model of the system’s nonlinear dynamics while taking into account the dynamic environment of the ocean. Most of these dynamics models do not incorporate data. The collection of data for AUVs is difficult, but necessary in order to have more flexibility in the model’s parameters due to the dynamic environment of the ocean. Yet, traditional system identification methods are still dominant today, despite the recent breakthroughs in Deep Learning. Therefore, in this thesis, we aim for a data-driven approach that embeds laws from physics in order to learn the state-space model of an AUV. More precisely, exploring the LNN framework for higher-dimensional systems. Furthermore, we also extend the LNN to account for non-conservative forces acting upon the system, such as damping and control inputs. The networks are trained to learn from simulated data of a second-order ordinary differential equation of an AUV. The trained model is evaluated by integrating paths from different initial states and comparing them to the true dynamics. The results yielded a model capable of predicting the output acceleration of the state space model but struggled in learning the direction of the forward movement with time. / I den här uppsatsen utforskas Lagrangianska Neurala Nätverk (LNN) för systemidentifiering av Autonoma Undervattensfordon (AUV) med 6 frihetsgrader. En av de största utmaningarna med AUV är deras begränsningar när det kommer till trådlös kommunikation och navigering under vatten. Ett krav för att ha fungerande AUV är deras förmåga att navigera och utföra uppdrag under okända undervattensförhållanden med begränsad och brusig sensordata. Dessutom är ett kritiskt krav för lokalisering och adaptiv reglerteknik att ha noggranna modeller av systemets olinjära dynamik, samtidigt som den dynamiska miljön i havet tas i beaktande. De flesta sådana modeller tar inte i beaktande sensordata för att reglera dess parameterar. Insamling av sådan data för AUVer är besvärligt, men nödvändigt för att skapa större flexibilitet hos modellens parametrar. Trots de senaste genombrotten inom djupinlärning är traditionella metoder av systemidentifiering dominanta än idag för AUV. Det är av dessa anledningar som vi i denna uppsats strävar efter en datadriven metod, där vi förankrar lagar från fysik under inlärningen av systemets state-space modell. Mer specifikt utforskar vi LNN för ett system med högre dimension. Vidare expanderar vi även LNN till att även ta ickekonservativa krafter som verkar på systemet i beaktande, såsom dämpning och styrsignaler. Nätverket tränas att lära sig från simulerad data från en andra ordningens differentialekvation som beskriver en AUV. Den tränade modellen utvärderas genom att iterativt integrera fram dess rörelse från olika initialstillstånd, vilket jämförs med den korrekta modellen. Resultaten visade en modell som till viss del var kapabel till att förutspå korrekt acceleration, med begränsad framgång i att lära sig korrekt rörelseriktning framåt i tiden.

The ocean symbolizes mystery, passion, and power. However, most of the ocean, about 80 %, is unknown to humans. AUVs (Autonomous Underwater Vehicles) provide a platform where terrain mapping, the biodiversity, and the resource survey of the ocean become accessible. Unlike ROVs (Remotely Operated Vehicle), AUVs operate according to their preset program which specifies the instructions required in different environments. One design aspects of AUVs that must be considered is that the data it acquire needs to be transmitted to a ground station (typically a ship). Although underwater acoustic communication is available nowadays, the low transmission rate and narrow bandwidth makes it unsuitable for large data transmission. For large sets of data, transmission with electromagnetic waves is more suitable. LoLo is an AUV which is designed and assembled at KTH Royal Institute of Technology, Sweden. Its wireless communication system consists of five components: RC (radio communication, 2.4 GHz), RF (radio frequency, 868 MHz), WIFI (wireless fidelity, 2.4 GHz), 4G (4th generation, 800 MHz, 1.8 GHz and 2.6 GHz) and GPS (global positioning system, 1.575 GHz). The goal of this project is to design an antenna board where the five subsystems are integrated. Importantly, due to the influence of seawater and waves, the resonant frequency of the antenna will fluctuate to a certain extent. Therefore, we need a robust, and preferably broadband, antenna system. In this project, we integrated printed dipole and monopole antennas on a single circuit board. The printed dipole antennas operate over a reasonable bandwidth and their radiation pattern is omnidirectional. The monopole antenna is designed to have multiple resonant frequencies which can cover BAND 20 (800 MHz) and BAND 3 (1.8 GHz) of the 4G service in Sweden. The 4G antenna shows good omnidirectional characteristics in the lower frequency band (band 20) and broadband characteristic in the higher frequency band. The upper 4G band is to be used to transmit large sets of data if a signal can be detected. The lower 4G band is added to provide redundancy. The antenna board is manufactured and measured. The results show the consistency with the simulation results and meets the requirement of the project. / Havet symboliserar mysterium, passion och kraft. Men det mesta av havet, cirka 80 %, är okänt för människor. AUVs (Autonomous Underwater Vehicles) är en plattform där terrängkartläggning, biologisk mångfald och resursundersökning blir tillgänglig. Till skillnad från ROVs (Remotely Operated Vehicles) fungerar AUVs enligt sitt förinställda program som specificerar de instruktioner som krävs i olika miljöer. Den data som den förvärvade måste överföras till en markstation (oftast en båt). Även om akustiska kommunikationen under vatten är möjlig idag gör den låga överföringshastigheten och den smala bandbredden den olämplig för stora dataöverföringar. I dessa fall är det bättre att överföra data med hjälp av elektromagnetiska vågor. LoLo är en AUV som är designad på KTH Royal Institute of Technology, Sverige. Dess trådlösa kommunikationssystem består av fem delsystem: RC (radiokommunikation, 2.4 GHz), RF (radiofrekvens, 868 GHz), WIFI (trådlös fidelity, 2.4 GHz), 4G (4 generationen av mobilnätverket, 800 MHz och 1.8 GHz) och GPS (global positioning system, 1.575 GHz). Målet med detta projekt är att designa antennerna för dessa fem delsystem. Viktigt att notera är antennernas resonansfrrekvens påverkas till viss del av havsvatten och vågor. Därför behövs vi ett robust, bredbandsantennsystem. I detta projekt integrerade vi dipolantenner och en monopolantenn på ett kretskort. Dipolantennerna har rimlig bandbredd och är omnidirektionella. Monopolantennen ger oss flera resonansfrekvenser som kan täcka Band 20 (800 MHz) och Band 3 (1.8 GHz) av 4Gspektrumet i Sverige. 4Gantennen visar omnidirektionella strålningsegenskaper i det lägre band et (band 20) och har vred bandbredd i det högre band et. Det högre bandet kommer användas för att skicka mycket data om en signal kan säkras. Det lägre bandet ger redundans. Antennen tillverkas och mäts i ett ekofritt rum. Mätresultaten stämmer överens med simuleringsresultaten och uppfyller projektets krav.

In year 2024 and 2025 the Royal Swedish Navy is expected to launch two new submarines with new possibilities to dock underwater vehicles. The submarines are part of the new Blekinge Class (A26) and will aid the Swedish Armed Forces and the Swedish Defense Materiel Administration (FMV) in their aim to develop and use more autonomous systems, to increase staff efficiency and to face the technological challenges of tomorrow. This thesis was carried out at FMV, with the purpose of investigating the physical requirements put on the new submarines, when docking an autonomous underwater vehicle. These requirements were identified through an analysis of qualitative and quantitative research. The analysis resulted in ten key insights, which led to thirteen requirements. The requirements were combined with project specific data of the AUV62 system and A26 submarine, to develop three conceptual designs of hard docking systems. The concepts were verified through analysis of material, stress and deflection, and geometric constraints. The concepts were evaluated based on how well they were fulfilling each requirement. A hammock-alike concept was shown to have most potential in being the most reliable hard docking system. The thesis ended with concluding that its purpose had been fulfilled, followed with recommendations for continued work. / Under 2024 och 2025 förväntas Svenska Marinen sjösätta två nya ubåtar, med nya förmågor gällande dockning av undervattensfordon. Ubåtarna ingår i nya Blekingeklass (A26) och är en del av Försvarsmaktens och Försvarets Materielverks (FMV) målsättning om att utveckla och använda mer autonoma system, för att öka personaleffektivitet och för att kunna möta morgondagens tekniska utmaningar. Detta examensarbete utfördes på uppdrag av FMV, med syftet att undersöka vilka fysiska krav som ställs för att hårddocka ett autonomt undervattensfordon på de nya ubåtarna. Dessa krav identifierades genom analys av en kvalitativ och kvantitativ undersökning. Analysen uppdagade tio nyckelinsikter som gav upphov till tretton krav. Kraven kombinerades med projektspecifik data för AUV62-systemet och ubåt A26, för att utveckla tre konceptuella designförslag av hårddockningssystem. Koncepten verifierades genom analys gällande material, spänning och utböjning, samt geometriska begränsningar. Koncepten utvärderades baserat på hur väl de uppfyllde respektive krav. Ett hängmatteformat koncept visade sig ha störst potential för att bli ett pålitligt hårddockningssystem. Examensarbetet avslutades med att projektets syfte ansågs vara uppfyllt, följt av förslag på vidare arbete.

<p dir="ltr">The technological advancements in marine robotics led to the expansion of the autonomous underwater vehicle (AUV) fleet. Depending on the applications, the type of the AUV ranges across various shapes and sizes. It seeks a solution for the issue of limited power capacity, often in terms of underwater docking systems. Underwater docking poses a significant challenge for AUVs, especially when considering the diverse shapes and sizes of these vehicles. Existing solutions usually are task specific, and do not address the idea of scalable underwater docking system design.<br>This thesis investigates the adaptability of the specific docking system design, previously validated for torpedo-shaped AUVs, to boxed-shaped AUVs in a nonlinear open water environment. In order to achieve this goal, the scalability of the docking system design of choice was tested in an open water non-linear underwater environment and validated. The scalability of the robust docking system was adapted to the box-shaped AUV, encompassing path planning, path following, and docking maneuver. The adapted docking system was based on the optic methods for docking station detection and subsequent docking. Additionally, the simulated environment was developed for the AUV model, for testing and debugging purposes. In the simulation, a custom PID controller was developed along with integrating the navigation and guidance package, to fully simulate the real life behavior of the AUV. </p><p dir="ltr">Furthermore, this work introduces a recurrent neural network-based architecture for investigating temporal dependencies of the sequential data input. The proposed architecture is based on CNN for spatial feature extraction and LSTM/GRU for temporal feature detection. The dataset collection is based on the simulation environment, by enhancing the artificial images with imposed realism. The dataset was gathered on different levels of turbidity and the collection process was automated.</p>