Abstract
Abstract Recently, research concerning the navigation of autonomous surface vehicles (ASVs) has been increasing. However, a large-scale implementation of these vessels is still held back by several challenges such as multi-object tracking. Attaining accurate object detection plays a big role in achieving successful tracking. This article presents the development of a detection model with an image-based Convolutional Neural Network trained through transfer learning, a deep learning technique. To train, test, and validate the detector module, data were collected with the SENSE ASV by sailing through two nearby ports, Leixões and Viana do Castelo, and recording video frames through its on-board cameras, along with a Light Detection And Ranging, GPS, and Inertial Measurement Unit data. Images were extracted from the collected data, composing a manually annotated dataset with nine classes of different vessels, along with data from other open-source maritime datasets. The developed model achieved a class mAP@[.5 .95] (mean average precision) of 89.5% and a clear improvement in boat detection compared to a multi-purposed state-of-the-art detector, YOLO-v4, with a 22.9% and 44.3% increase in the mAP with an Intersection over Union threshold of 50% and the mAP@[.5 .95], respectively. It was integrated in a detection and tracking system, being able to continuously detect nearby vessels and provide sufficient information for simple navigation tasks.

Abstract
Structures associated with offshore wind energy production require an arduous and cyclical inspection and maintenance (O&M) procedure. Moreover, the harsh challenges introduced by sub-sea phenomena hamper visibility, considerably affecting underwater missions. The lack of quality 3D information within these environments hinders the applicability of autonomous solutions in close-range navigation, fault inspection and intervention tasks since these have a very poor perception of the surrounding space. Deep learning techniques are widely used to solve these challenges in aerial scenarios. The developments in this subject are limited regarding underwater environments due to the lack of publicly disseminated underwater information. This article presents a new underwater dataset: NEREON, containing both 2D and 3D data gathered within real underwater environments at the ATLANTIS Coastal Test Centre. This dataset is adequate for monocular depth estimation tasks, which can provide useful information during O&M missions. With this in mind, a benchmark comparing different deep learning approaches in the literature was conducted and presented along with the NEREON dataset.

Abstract
Given the increased interest in offshore wind energy, there is a greater need for advancements in operation and maintenance technology. As a result, robotic solutions are required to avoid human risky behavior and reduce associated operational costs. In order to accommodate the need for inspecting multiple domains, multiple robotic vehicles are utilized, which requires the deployment of control stations that can effectively monitor, facilitate communication among different vehicles, and ensure successful completion of the overall mission. A shore control centre (SCC) is a communication software infrastructure capable of monitoring, localizing and planning missions for a group of multi-domain heterogeneous robots within a local network. This paper proposes an SCC as: (i) an active monitor by continuously observing the local behaviour of each robot and the global progress of the mission and its safety; (ii) a mission planner that provides and supervises its execution while constantly checking for critical failures and intervening in the case of unexpected events. Also, The control centre is able to connect to multiple vehicles from various domains and monitor real-time data. Accordingly, validation procedures were carried out in real conditions.

Abstract
<p>An intelligible step-by-step Reinforcement Learning (RL) problem formulation and the availability of an easy-to-use demonstrative toolbox for students at various levels (e.g., undergraduate, bachelor, master, doctorate), researchers and educators. This tool facilitates the familiarization with the key concepts of RL, its problem formulation and implementation. The results demonstrated in this paper are produced by a Python program that is released open-source, along with other lecture materials to reduce the learning barriers in such innovative research topic in robotics.</p> <p>The RL paradigm is showing promising results as a generic purpose framework for solving decision-making problems (e.g., robotics, games, finance). In this work, RL is used for solving a robotics 2D navigational problem where the robot needs to avoid collisions with obstacles while aiming to reach a goal point. A navigational problem is simple and convenient for educational purposes, since the outcome is unambiguous (e.g., the goal is reached or not, a collision happened or not). Thus, the intent is to accelerate the adoption of RL techniques in the field of mobile robotics.</p> <p>Motivate and promote the adoption of RL techniques to solve decision-making problems, specifically in robotics. </p> <p>Due to a lack of accessible educational and demonstrative toolboxes concerning the field of RL, this work combines theoretical exposition with an accessible open-source graphical interactive toolbox to facilitate the apprehension.</p> <p>This study aims to reduce the learning barriers and inspire young students, researchers and educators to use RL as an obvious tool to solve robotics problems.</p>

Abstract
In the present day, unmanned aerial vehicle (UAV) technology is being used for a multitude of inspection operations, including those in offshore structures such as wind-farms. Due to the distance of these structures to the coast, drones need to be carried to these structures via ship. To achieve a completely autonomous operation, the UAV can greatly benefit from an autonomous surface vehicle (ASV) to transport the UAV to the operation location and coordinate a successful landing between the two. This work presents the concept of a four-link parallel platform to perform wave-motion synchronization to facilitate UAV landings. The parallel platform consists of two base floaters connected with rigid rods, linked by linear actuators to a top mobile platform for the landing of a UAV. Using an inverse kinematics approach, a study of the position of the cylinders for greater range of motion and a workspace analysis is achieved. The platform makes use of a feedback controller to reduce the total motion of the landing platform. Using the robotic operating system (ROS) and Gazebo to emulate wave motions and represent the physical model and actuator system, the platform control system was successfully validated.

Abstract
Exploiting stronger winds at offshore farms leads to a cyclical need for maintenance due to the harsh maritime conditions. While autonomous vehicles are the prone solution for O&M procedures, sub-sea phenomena induce severe data degradation that hinders the vessel's 3D perception. This article demonstrates a hybrid underwater imaging system that is capable of retrieving tri-dimensional information: dense and textured Photogrammetric Stereo (PS) point clouds and multiple accurate sets of points through Light Stripe Ranging (LSR), that are combined into a single dense and accurate representation. Two novel fusion algorithms are introduced in this manuscript. A Joint Masked Regression (JMR) methodology propagates sparse LSR information towards the PS point cloud, exploiting homogeneous regions around each beam projection. Regression curves then correlate depth readings from both inputs to correct the stereo-based information. On the other hand, the learning-based solution (RHEA) follows an early-fusion approach where features are conjointly learned from a coupled representation of both 3D inputs. A synthetic-to-real training scheme is employed to bypass domain-adaptation stages, enabling direct deployment in underwater contexts. Evaluation is conducted through extensive trials in simulation, controlled underwater environments, and within a real application at the ATLANTIS Coastal Testbed. Both methods estimate improved output point clouds, with RHEA achieving an average RMSE of 0.0097 m -a 52.45% improvement when compared to the PS input. Performance with real underwater information proves that RHEA is robust in dealing with degraded input information; JMR is more affected by missing information, excelling when the LSR data provides a complete representation of the scenario, and struggling otherwise.