Autonomous Sailboat Research Articles

Simulation-based Reinforcement and Imitation Learning for Autonomous Sailboat Navigation in Variable Environmental Conditions

Simulation-based Reinforcement and Imitation Learning for Autonomous Sailboat Navigation in Variable Environmental Conditions

Physics Applied to the Sailboat Optimum Upwind Path Algorithm

Autonomous vehicles are predicted to be an enormous disrupter to our economy, creating a massive and unpredictable impact, much like the smartphone on telecommunications. Marine transportation is at the leading edge, with automated boats crossing the oceans. Sailboats and sail-assisted boats, in particular, are not just efficient and environmentally friendly, but also allow automated boats to be at sea for extended periods. This is excellent for applications such as oil and gas pipeline monitoring, data collection, and ocean mapping. Navigation is critical, however, one issue unique to sailboats is that it is impossible to sail directly into the wind. In order to navigate upwind the boat needs to follow a zig-zag route towards the destination. Details such as the angles, lengths, and number of zig-zags are critical to maximizing forward progress to the destination. This research includes an approach to determine the optimal zig-zag route for efficiency, which could be used as part of route planning before a journey begins, or as part of an autonomous sailboat system.

Complete holography-based system for the identification of microparticles in water samples.

Here, we present a comprehensive holography-based system designed for detecting microparticles through microscopic holographic projections of water samples. This system is designed for researchers who may be unfamiliar with holographic technology but are engaged in microparticle research, particularly in the field of water analysis. Additionally, our innovative system can be deployed for environmental monitoring as a component of an autonomous sailboat robot. Our system's primary application is for large-scale classification of diverse microplastics that are prevalent in water bodies worldwide. This paper provides a step-by-step guide for constructing our system and outlines its entire processing pipeline, including hologram acquisition for image reconstruction.

A Hardware-in-the-Loop Simulator to Optimize Autonomous Sailboat Performance in Real Ocean Conditions

In this work, a hardware-in-the-loop (HIL) simulator is designed to diagnose the behavior of an autonomous sailboat as it navigates between waypoints. At its core, the HIL simulator includes the sailboat pilot on an embedded system. The sensor data input to the embedded system is fed by a navigation simulator that takes into account the different forces on the sailboat due to the wind, waves and current conditions. The HIL simulator is then tested for a navigation route from sea trials published in 2014, and the behavior of the automated pilot is compared to its behavior when the vessel is driven by a crew. As demonstrated, the automated system can outperform the man-operated vessel. The tool is also used to diagnose weaknesses in the sailboat autopilot algorithm that can be improved in the future.

V-Stability Based Control for Energy-Saving Towards Long Range Sailing

Ensuring an adequate supply of energy is a challenge in the development of autonomous sailboats for long-range sailing capability, as the continuous high-frequency execution of control commands leads to high energy consumption. However, as the control of sailboat is much more complex than the other USVs, especially in a complicated and windy environment, simply reducing control frequency will lead to large path tracking error, and even unsuccessful navigation. Based on the V-stability controller of sailboats, this letter proposes an energy-saving (E-saving) control scheme for a sailboat, to enable the energy planning of actuators. While keeping the sailboat traveling within the acceptable range along a desired path, we develop an approach to analyze and reach a satisfying tradeoff between path tracking error and consumed energy. The E-saving method is examined in simulated environments and field experiments using an autonomous sailboat (OceanVoy). The results show that E-saving method reduces energy consumption by approximately 11% compared to that of the previous V-stability controller, indicating that our method can elongate sailing distance.

Comparison of Reinforcement and Imitation Learning algorithms in autonomous sailboat Digital Twins

This project aims to study the performance of two reinforcement machine learning algorithms, namely the Proximal Policy Optimization and Soft Actor Critic, in the simulation of autonomous sailboats and their response to different wind directions while avoiding obstacles detected by image analysis and following defined target check-points. Also, the effect of the imitation learning algorithms Behavioral Cloning and Generative Adversarial Imitation Learning combined with the first mentioned algorithms is studied. The proposed scenarios consist of areas filled with random static or moving obstacles and with the presence of favorable or crosswinds. The motivation for the project comes from the lack of studies of the mentioned algorithms in autonomous sailboats, issue which the current study tries to address. The Unity platform and ML-Agents machine learning toolkit are used for development and the methodology that guides the project can be similarly applied to other reinforcement learning problems. Through agent training, it is possible to compare the results and observe that the Proximal Policy Optimization obtains better performance within the proposed scenarios, both with and without the support of imitation learning algorithms.

Influence of Autonomous Sailboat Dual-Wing Sail Interaction on Lift Coefficients

Influence of Autonomous Sailboat Dual-Wing Sail Interaction on Lift Coefficients

Sustainable Solutions for Sea Monitoring With Robotic Sailboats: N-Boat and F-Boat Twins.

Strategic management and production of internal energy in autonomous robots is becoming a research topic with growing importance, especially for platforms that target long-endurance missions, with long-range and duration. It is fundamental for autonomous vehicles to have energy self-generation capability to improve energy autonomy, especially in situations where refueling is not viable, such as an autonomous sailboat in ocean traversing. Hence, the development of energy estimation and management solutions is an important research topic to better optimize the use of available energy supply and generation potential. In this work, we revisit the challenges behind the project design and construction for two fully autonomous sailboats and propose a methodology based on the Restricted Boltzmann Machine (RBM) in order to find the best way to manage the supplementary energy generated by solar panels. To verify the approach, we introduce a case study with our two developed sailboats that have planned payload with electric and electronics, and one of them is equipped with an electrical engine that may eventually help with the sailboat propulsion. Our current results show that it is possible to augment the system confidence level for the potential energy that can be harvested from the environment and the remaining energy stored, optimizing the energy usage of autonomous vehicles and improving their energy robustness.

Evaluation and development of wireless monitoring systems for an autonomous sailboat

In this study, the performance of the Long-Range (LoRa) communication and 3DR Radio Telemetry modules was evaluated. The suitability of these two communication options was assessed for the application on an autonomous sailboat prototype. The communication range and power consumption of each module are the primary factors for selection. The test results found that LoRa technology can send data stably with a range of up to 10.5 km, while the 3DR Radio Telemetry module can only send data stably up to 300 m. The LoRa technology consumes 9.56 Wh of electrical energy per day for the autonomous sailboat application, while the 3DR Radio Telemetry consumes 11.07 Wh of electrical energy per day. It was decided that LoRA technology would be implemented on the autonomous sailboat.

An Engineering Design Approach for the Development of an Autonomous Sailboat to Cross the Atlantic Ocean

Over the past decade, interest in autonomous vessels significantly increased as the technology improved, especially in the automotive industry. Unlike cars, ships travel in a wild environment and maritime lanes are not limited by white lines. This makes the design of fully autonomous vessels even more challenging. Additionally, the need to reduce greenhouse gas emissions led to a renewed interest in wind propulsion. Sailboats have several advantages, such as full energy autonomy and a limited environmental impact. The Microtransat Challenge, which consists of crossing the Atlantic Ocean, is a tremendous test field. This paper describes, within that frame, a design procedure for the development of a robust fully autonomous sailboat to be deployed for long-term missions. In this paper, the mechanical and electronic design strategies are presented. A focus is on reliability and power management. Moreover, a test procedure for validating each design increment is described as well as a path plan that considers the risk of collision and weather routing with wind and currents. The Microtransat remains a challenge that no autonomous ship has ever succeeded (and has been completed by a single unmanned vessel, SB Met in 2018). However, the results by Breizh Tigresse and Sealeon in 2015 and 2018 made a step forward in terms of time and distance. They are presented and analyzed in this work.

Generic and Flexible Unmanned Sailboat for Innovative Education and World Robotic Sailing Championship.

Over the past two decades, scholars developed various unmanned sailboat platforms, but most of them have specialized designs and controllers. Whereas these robotic sailboats have good performance with open-source designs, it is actually hard for interested researchers or fans to follow and make their own sailboats with these open-source designs. Thus, in this paper, a generic and flexible unmanned sailboat platform with easy access to the hardware and software architectures is designed and tested. The commonly used 1-m class RC racing sailboat was employed to install Pixhawk V2.4.8, Arduino Mega 2,560, GPS module M8N, custom-designed wind direction sensor, and wireless 433 Mhz telegram. The widely used open-source hardware modules were selected to keep reliable and low-cost hardware setup to emphasize the generality and feasibility of the unmanned sailboat platform. In software architecture, the Pixhawk V2.4.8 provided reliable states’ feedback. The Arduino Mega 2,560 received estimated states from Pixhawk V2.4.8 and the wind vane sensor, and then controlled servo actuators of rudder and sail using simplified algorithms. Due to the complexity of introducing robot operating system and its packages, we designed a generic but real-time software architecture just using Arduino Mega 2,560. A suitable line-of-sight guidance strategy and PID-based controllers were used to let the autonomous sailboat sail at user-defined waypoints. Field tests validated the sailing performance in facing WRSC challenges. Results of fleet race, station keeping, and area scanning proved that our design and algorithms could control the 1-m class RC sailboat with acceptable accuracy. The proposed design and algorithms contributed to developing educational, low-cost, micro class autonomous sailboats with accessible, generic, and flexible hardware and software. Besides, our sailboat platform also facilitates readers to develop similar sailboats with more focus on their missions.

High-Level Path Planning for an Autonomous Sailboat Robot Using Q-Learning.

Path planning for sailboat robots is a challenging task particularly due to the kinematics and dynamics modelling of such kinds of wind propelled boats. The problem is divided into two layers. The first one is global where a general trajectory composed of waypoints is planned, which can be done automatically based on some variables such as weather conditions or defined by hand using some human–robot interface (a ground-station). In the second local layer, at execution time, the global route should be followed by making the sailboat proceed between each pair of consecutive waypoints. Our proposal in this paper is an algorithm for the global, path generation layer, which has been developed for the N-Boat (The Sailboat Robot project), in order to compute feasible sailing routes between a start and a target point while avoiding dangerous situations such as obstacles and borders. A reinforcement learning approach (Q-Learning) is used based on a reward matrix and a set of actions that changes according to wind directions to account for the dead zone, which is the region against the wind where the sailboat can not gain velocity. Our algorithm generates straight and zigzag paths accounting for wind direction. The path generated also guarantees the sailboat safety and robustness, enabling it to sail for long periods of time, depending only on the start and target points defined for this global planning. The result is the development of a complete path planner algorithm that, together with the local planner solved in previous work, can be used to allow the final developments of an N-Boat making it a fully autonomous sailboat.

Analysis of Influencing Factors on Lift Coefficients of Autonomous Sailboat Double Sail Propulsion System Based on Vortex Panel Method

Sail is the core part of autonomous sailboat and wing sail is a new type of sail. Wing sail generates not only propulsion but also lateral force and heeling moment. The latter two will affect the navigation status and bring resistance. Double sail can effectively reduce the center of wind pressure and heeling moment. In order to study the effect of distance between two sails, airfoil and attack angle on the total lift coefficient of double sail propulsion system, pressure coefficient distribution and lift coefficient calculation model have been established based on vortex panel method. By using the basic finite solution, the fluid dynamic forces on the two-dimensional sails are computed. The results show that, the distance in the range of 0 to 1 time chord length, when using the same airfoil in the fore and aft sail, the total lift coefficient of the double sail increases with the increase of distance, finally reaches a stable value in the range of one to three times chord length. Lift coefficients of thicker airfoils are more sensitive to the change of distance. The thicker the airfoil, the longer distance is required of the total lift coefficient toward stable. When different airfoils are adopted in fore and aft sail, the total lift coefficient increases with the increase of the thickness of aft sail. The smaller the thickness difference is, the more sensitive to the distance change the lift coefficient is. The thinner the fore sail is, the lower the influence will be on the lift coefficient of aft sail.

Coastal Harmful Algae Bloom Monitoring via a Sustainable, Sail-Powered Mobile Platform

Harmful algae blooms (HAB) in coastal marine environments are increasing in number and duration, pressuring local resource managers to implement mitigation solutions to protect human and ecosystem health. However, insufficient spatial and temporal observations create uninformed management decisions. In order to better detect and map blooms, as well as the environmental conditions responsible for their formation, long-term, unattended observation platforms are desired. In this article, we describe a new cost-efficient, autonomous, mobile platform capable of accepting several sensors that can be used to monitor harmful algae blooms in near real-time. The Navocean autonomous sail-powered surface vehicle is deployable by a single person from shore, capable of waypoint navigation in shallow and deep waters, and powered completely by renewable energy. We present results from three surveys of the Florida Red Tide harmful algae bloom (Karenia brevis) of 2017-2018. The vessel made significant progress towards waypoints regardless of wind conditions while underway measurements revealed patches of elevated chl. a likely attributable to the K. brevis blooms as based on ancillary measurements. Measurements of colored dissolved organic matter (CDOM) and turbidity provided an environmental context for the blooms. While the autonomous sailboat directly adds to our phytoplankton/HAB monitoring capabilities, the package may also help to ground-truth satellite measurements of HABs if careful validation measurements are performed. Finally, several other pending and future use cases for coastal and inland monitoring are discussed. To our knowledge, this is the first demonstration of a sail-driven vessel used for coastal HAB monitoring.

Autopilot Design for an Autonomous Sailboat Based on Sliding Mode Control

This paper deals with the design of an autopilot based on sliding mode control combined with feedback linearization method for an autonomous sailing vessel. This autopilot is developed using a model with four degrees of freedom, which represents the dynamic of a sailboat. Due to the high nonlinearity of the developed dynamic model, heading and sail opening angle controller were developed to steer the sailboat toward a specific target position. The nonlinear four degrees of freedom dynamic model for the sailing vessel is first described. Then, an autopilot is designed using both sliding mode control and feedback linearization method. Finally, some simulations are carried out to illustrate the behavior of the overall system.

Adaptive Pid Control Of An Autonomous Sailboat

This paper discusses the control architecture for autonomous sailboat navigation. The control of non linear autonomous sailboat system is done by using Conventional PID controller (CPID) and Adaptive PID controller (APID). This mainly focuses on the way of dynamically controlling the sail of the sailboat to reduce the heel angle and roll motion caused by the wind. A large heeling angle, a lot of roll motion due to wind gusts can cause the boat to capsize. Sailboats cannot sail directly up-wind either and one have to move in a zigzag pattern to be able to sail up-wind. The sail angle is optimized and rudder angle is controlled for the given wind direction and desired moment. The main goal is to allow long endurance autonomous mission.

Position keeping control of an autonomous sailboat

This paper addresses the problem of reaching and keeping a target position for a sailboat. A method to control sail angle is proposed, using an adaptive adjustment of the sail to regulate the acceleration of the sailboat. A tacking strategy is defined to navigate into the wind, arrive upwind to slow down the boat and then to stay close to the target point. Simulation results show the effectiveness of the proposed approach. The stability of the sailboat using the proposed control has been proven.

Integrated Design and Testing of an Anemometer for Autonomous Sail Drones

A correct estimation of both direction and intensity of wind velocity is fundamental for controlling an autonomous sail-boat. This kind of estimation has to be performed in a harsh environment considering the direct exposition of the sensor to salt, fog, and to any variable weather conditions. An important feature is represented by the sensor size, which has to be small compared to the drone size. Costs have to be optimized with respect to the overall small budget involved in the construction of the drone. Finally, extensive use on drones or in large sensor networks should be greatly advantaged by an easy substitutability in the case of accidental damage or system loss, an eventuality which is difficult to be completely avoided for large scale, prolonged monitoring activities. In this work authors propose a low cost ultrasonic planar anemometer with a very interesting price to performance ratio which is obtained by introducing a simple, original and innovative Arduino based architecture. Preliminary design and the results of calibration will be described, followed by testing activities performed on a low-speed large section wind tunnel, available at University of Florence supported by simple but effective computational fluid dynamic (CFD) simulations.

Towards a Real-Time Embedded System for Water Monitoring Installed in a Robotic Sailboat.

Problems related to quality (and quantity) of water in natural resources or in artificial reservoirs are frequently arising and are at the center of attention of authorities and governments around the world. Many times the monitoring is not performed in an efficient time frame and a precise manner, whereas the adoption of fast and punctual solutions would undoubtedly improve the water quality and consequently enhance the life of people. To minimize or diminish such kinds of problems, we propose an architecture for sensors installed in a robotic platform, an autonomous sail boat, able to acquire raw data relative to water quality, to process and make them available to people that might be interested in such information. The main contributions are the sensors architecture itself, which uses low cost sensors, with practical experimentation done with a prototype. Results show data collected for points in lakes and rivers in the northeast of Brazil. This embedded system is fixed in the sailboat robot with the intention to facilitate the study of water quality for long endurance missions. This robot can help monitoring water bodies in a more consistent manner. Nonetheless the system can also be used with fixed vases or buoys in strategic points.

Data-Driven Online Speed Optimization in Autonomous Sailboats

This paper addresses the issue of data-driven online velocity optimization of an autonomous sailboat. Autonomous sailboats represent an ideal for long range and duration reconnaissance missions. Sailboat control is a challenging control task; sailboats are characterized by a number of control variables, all of which affect the ship trajectory and state in a highly nonlinear fashion. In this paper, a path-following automatic sailboat controller is presented. The control system has two main components: a heading control, acting on the rudder, and a velocity optimizer, acting on the sails. The optimizer is based on a modified extremum seeking approach. This paper also derives a first-principle-based 4 DoF sailboat model that is experimentally validated and used to guide the design and tuning of the control system. In fact, the control system is first tuned and validated in simulation. The simulation environment enables the comparison of the proposed model against a theoretical benchmark and a state-of-the-art controller. The analysis reveals that the proposed control system achieves near-optimal performance and considerably outperforms the state-of-the-art solution. Finally, the controller is tested and validated on an instrumented scale model.