Shared autonomy for low‐cost underwater vehicles

This field report presents an overview of the development and testing of a semi‐autonomous underwater vehicle (sAUV). The work presented here is aimed at bridging the gap between current remotely operated vehicles and autonomous research platforms by developing shared autonomy capabilities for low‐cost underwater vehicles. We use commercially available components and open‐source software interfaces to provide a wider range of capabilities for underwater autonomy research at a lower cost than previously available systems. We describe the overall structure of the system, discuss its capabilities, and provide results demonstrating system performance. We place particular emphasis on shared autonomy, where a human operator is assisted in controlling an underwater tethered vehicle. We present three capabilities developed for the sAUV: (a) an assisted control mode that provides a variable level of assistance using an on‐line estimate of user skill level, (b) a planner to generate paths that avoid tether entanglement, and (c) a sonar processing algorithm that identifies informative sonar images for selecting features for 3D scene reconstruction. The vehicle has been deployed on five off‐shore and near‐shore marine field deployments since 2015, and this report includes selected results from four of those trials to demonstrate the capabilities and limitations of the sAUV system.     

The Malta cistern mapping project: Underwater robot mapping and localization within ancient tunnel systems

This paper documents the development of an underwater robot system enabled with several mapping and localization techniques applied to a particular archaeological expedition. The goal of the expedition was to explore and map ancient cisterns located on the islands of Malta and Gozo. The cisterns of interest acted as water storage systems for fortresses, private homes, and churches. Such cisterns often consisted of several connected chambers, still containing water. A sonar‐equipped remotely operated vehicle (ROV) was deployed into these cisterns to obtain both video footage and sonar range measurements. Six different mapping and localization techniques were employed, including (1) sonar image mosaics using stationary sonar scans, (2) sonar image mosaics using stationary sonar scans with Smart Tether position data, (3) simultaneous localization and mapping (SLAM) while the vehicle was in motion, (4) SLAM using stationary sonar scans, (5) localization using previously created maps, and (6) SLAM while the vehicle was in motion with Smart Tether position data. Top‐down‐view maps of 22 different cisterns were successfully constructed. It is estimated that the cisterns were built as far back as 300 B.C., and few records of their size, shape, and connectivity existed before the expedition. © 2010 Wiley Periodicals, Inc.     

Three‐dimensional reconstruction of underwater objects using wide‐aperture imaging SONAR

The estimation of the geometric structure of objects located underwater underpins a plethora of applications such as mapping shipwrecks for archaeology, monitoring the health of coral reefs, detecting faults in offshore oil rigs and pipelines, detection and identification of potential threats on the seabed, etc. Acoustic imaging is the most popular choice for underwater sensing. Underwater exploratory vehicles typically employ wide‐aperture Sound Navigation and Ranging (SONAR) imaging sensors. Although their wide aperture enables scouring large volumes of water ahead of them for obstacles, the resulting images produced are blurry due to integration over the aperture. Performing three‐dimensional (3D) reconstruction from this blurry data is notoriously difficult. This challenging inverse problem is further exacerbated by the presence of speckle noise and reverberations. The state‐of‐the‐art methods in 3D reconstruction from sonar either require bulky and expensive matrix‐arrays of sonar sensors or additional narrow‐aperture sensors. Due to its low footprint, the latter induces gaps between reconstructed scans. Avoiding such gaps requires slow and cumbersome scanning by the vehicles that carry the scanners. In this paper, we present two reconstruction methods enabling on‐site 3D reconstruction from imaging sonars of any aperture. The first of these presents an elegant linear formulation of the problem, as a blind deconvolution with a spatially varying kernel. The second method is a simple algorithmic approach for approximate reconstruction, using a nonlinear formulation. We demonstrate that our simple approximation algorithms perform 3D reconstruction directly from the data recorded by wide‐aperture systems, thus eliminating the need for multiple sensors to be mounted on underwater vehicles for this purpose. Additionally, we observe that the wide aperture may be exploited to improve the coverage of the reconstructed samples (on the scanned object's surface). We demonstrate the efficacy of our algorithms on simulated as well as real data acquired using two sensors, and we compare our work to the state of the art in sonar reconstruction. Finally, we show the employability of our reconstruction methods on field data gathered by an autonomous underwater vehicle.     

Teach‐and‐repeat path following for an autonomous underwater vehicle

This paper presents a teach‐and‐repeat path‐following method for an autonomous underwater vehicle (AUV) navigating long distances in environments where external navigation aides are denied. This method utilizes sonar images to construct a series of reference views along a path, stored as a topological map. The AUV can then renavigate along this path, either to return to the start location or to repeat the route. Utilizing unique assumptions about the sonar image‐generation process, this system exhibits robust image‐matching capabilities, providing observations to a discrete Bayesian filter that maintains an estimate of progress along the path. Image‐matching also provides an estimate of offset from the path, allowing the AUV to correct its heading and effectively close the gap. Over a series of field trials, this system demonstrated online control of an AUV in the ocean environment of Holyrood Arm, Newfoundland and Labrador, Canada. The system was implemented on an International Submarine Engineering Ltd. Explorer AUV and performed multiple path completions over both a 1 and 5 km track. These trials illustrated an AUV operating in a fully autonomous mode, in which navigation was driven solely by sensor feedback and adaptive control. Path‐following performance was as desired, with the AUV maintaining close offset to the path.     

Multi‐AUV motion planning for archeological site mapping and photogrammetric reconstruction

This paper presents coupled and decoupled multi‐autonomous underwater vehicle (AUV) motion planning approaches for maximizing information gain. The work is motivated by applications in which multiple AUVs are tasked with obtaining video footage for the photogrammetric reconstruction of underwater archeological sites. Each AUV is equipped with a video camera and side‐scan sonar. The side‐scan sonar is used to initially collect low‐resolution data to construct an information map of the site. Coupled and decoupled motion planning approaches with respect to this map are presented. Both planning methods seek to generate multi‐AUV trajectories that capture close‐up video footage of a site from a variety of different viewpoints, building on prior work in single‐AUV rapidly exploring random tree (RRT) motion planning. The coupled and decoupled planners are compared in simulation. In addition, the multiple AUV trajectories constructed by each planner were executed at archeological sites located off the coast of Malta, albeit by a single‐AUV due to limited resources. Specifically, each AUV trajectory for a plan was executed in sequence instead of simultaneously. Modifications are also made by both planners to a baseline RRT algorithm. The results of the paper present a number of trade‐offs between the two planning approaches and demonstrate a large improvement in map coverage efficiency and runtime.     

In‐water visual ship hull inspection using a hover‐capable underwater vehicle with stereo vision

Underwater visual inspection is an important task for checking the structural integrity and biofouling of the ship hull surface to improve the operational safety and efficiency of ships and floating vessels. This paper describes the development of an autonomous in‐water visual inspection system and its application to visual hull inspection of a full‐scale ship. The developed system includes a hardware vehicle platform and software algorithms for autonomous operation of the vehicle. The algorithms for vehicle autonomy consist of the guidance, navigation, and control algorithms for real‐time and onboard operation of the vehicle around the hull surface. The environmental perception of the developed system is mainly based on optical camera images, and various computer vision and optimization algorithms are used for vision‐based navigation and visual mapping. In particular, a stereo camera is installed on the underwater vehicle to estimate instantaneous surface normal vectors, which enables high‐precision navigation and robust visual mapping, not only on flat areas but also over moderately curved hull surface areas. The development process of the vehicle platform and the implemented algorithms are described. The results of the field experiment with a full‐scale ship in a real sea environment are presented to demonstrate the feasibility and practical performance of the developed system.     

Map Building Fusing Acoustic and Visual Information using Autonomous Underwater Vehicles

We present a system for automatically building three‐dimensional (3‐D) maps of underwater terrain fusing visual data from a single camera with range data from multibeam sonar. The six‐degree‐of‐freedom location of the camera relative to the navigation frame is derived as part of the mapping process, as are the attitude offsets of the multibeam head and the onboard velocity sensor. The system uses pose graph optimization and the square root information smoothing and mapping framework to simultaneously solve for the robot's trajectory, the map, and the camera location in the robot's frame. Matched visual features are treated within the pose graph as images of 3‐D landmarks, while multibeam bathymetry submap matches are used to impose relative pose constraints linking robot poses from distinct tracklines of the dive trajectory. The navigation and mapping system presented works under a variety of deployment scenarios on robots with diverse sensor suites. The results of using the system to map the structure and the appearance of a section of coral reef are presented using data acquired by the Seabed autonomous underwater vehicle.     

Toward Autonomous Exploration in Confined Underwater Environments

In this field note, we detail the operations and discuss the results of an experiment conducted in the unstructured environment of an underwater cave complex using an autonomous underwater vehicle (AUV). For this experiment, the AUV was equipped with two acoustic sonar sensors to simultaneously map the caves' horizontal and vertical surfaces. Although the caves' spatial complexity required AUV guidance by a diver, this field deployment successfully demonstrates a scan‐matching algorithm in a simultaneous localization and mapping framework that significantly reduces and bounds the localization error for fully autonomous navigation. These methods are generalizable for AUV exploration in confined underwater environments where surfacing or predeployment of localization equipment is not feasible, and they may provide a useful step toward AUV utilization as a response tool in confined underwater disaster areas.     

Underwater confined space mapping by resource‐constrained autonomous vehicle

For marine industrial inspection, archaeology, and geological formation study, the ability to map unknown underwater enclosed and confined spaces is desirable and well suited for robotic vehicles. To date, there are few solutions thoroughly tested in the field designed to perform this specific task, none of which operate autonomously. With a small, low‐cost biomimetic platform known as the U‐CAT, we developed a mapping‐mission software architecture in which the vehicle executes three key sensor‐based reactive stages: entering, exploring, and exiting. Encapsulated in the exploring stage are several state‐defined navigation strategies, called patterns, which were designed and initially tested in simulation. The results of simulation work informed the selection of two patterns that were executed in field trials at a submerged building in Rummu Quarry Lake, Estonia, as part of several full mapping missions. Over the course of these trials, the vehicle was capable of observing the majority (78–97%) of 49.9 explorable square meters within 7 minutes. Based on these results, we demonstrate the capability of a low‐cost and resource‐constrained vehicle to perform confined space mapping under sensor uncertainty. Further, the observations made by the vehicle are shown to be suitable for a target site reconstruction and analysis in postprocessing, which is the intended outcome of this type of mission in practical applications.     

一种海洋温度和盐度的虚拟生成方法

海洋温度和盐度精确仿真的主要难点是实际实验数据的获得成本高且操作不便,根据有限的实验数据对温度和盐度进行仿真,对研究它们对水下机器人声纳系统及载体控制系统的影响具有重要意义。针对UUV多水下机器人的数字仿真平台的具体工程提出了一种海洋温度和盐度的虚拟生成方法。对海洋中海水的温度和盐度分别进行仿真:水平面采用二元全区间插值方法,垂直面的仿真分为两种情况:对于已有实验数据深度范围内的温度和盐度数值采用三次样条插值,而对于未知实验数据深度范围内的温度和盐度数值采用曲线拟合的方法来近似获得,方法能适用于普遍海域的温度和盐度的虚拟生成,为研究海洋温度和盐度对海洋生物或水下工程的研究提供参考。     

Magnetic survey and autonomous target reacquisition with a scalar magnetometer on a small AUV

A scalar magnetometer payload has been developed and integrated into a two‐man portable autonomous underwater vehicle (AUV) for geophysical and archeological surveys. The compact system collects data from a Geometrics microfabricated atomic magnetometer, a total‐field atomic magnetometer. Data from the sensor is both stored for post‐processing and made available to an onboard autonomy engine for real‐time sense and react behaviors. This system has been characterized both in controlled laboratory conditions and at sea to determine its performance limits. Methodologies for processing the magnetometer data to correct for interference and error introduced by the AUV platform were developed to improve sensing performance. When conducting seabed surveys, detection and characterization of targets of interest are performed in real‐time aboard the AUV. This system is used to drive both single‐ and multiple‐vehicle autonomous target reacquisition behaviors. The combination of on‐board target detection and autonomous reacquire capability is found to increase the effective survey coverage rate of the AUV‐based magnetic sensing system.     

水下无人装备声呐发射机自动测试系统设计

针对某型水下无人装备声呐发射机的工作原理和现有声呐发射机检测方式落后,检测能力不足等问题,设计了基于国产PXI总线仪器平台的水下无人装备声呐发射机自动测试系统;首先分析了发射机的工作原理和检测需求,进行声呐发射机自动测试系统的总体设计;其次介绍自动测试系统的软硬件平台实现方法和关键技术,包括国产PXI总线仪器选型、接口单元设计、测试平台软件架构设计和实现以及数据处理算法设计和实现等技术;最后对声呐发射机自动测试系统的试验结果进行分析;应用结果表明：该系统能实现某型水下无人装备声呐发射机主要参数的测量,相较传统测试方法,自动测试系统能够节约大量的测试时间与人力,实现测试工作效果的最大化提高,并且还能防止人为因素对测试结果造成的不良影响。     

Loon Copter: Implementation of a hybrid unmanned aquatic–aerial quadcopter with active buoyancy control

Aquatic–aerial unmanned vehicles recently became the focus of many researchers due to their various possible applications. Achieving a fully operational vehicle that is capable of aerial, water‐surface, and underwater operations is a significant challenge considering the vehicle's air–water–air transition, propulsion system, and stability underwater. We present in this paper an unconventional unmanned hybrid aquatic–aerial quadcopter with active buoyancy control that is capable of aerial flight and water‐surface operation, as well as subaquatic diving. We report on the first successful prototype of the vehicle, named the Loon Copter, to provide initial evaluation results of its performance in both mediums. The Loon Copter uses a single set of motors and propellers for both air and underwater maneuvering. It utilizes a ballast system to control vehicle buoyancy and depth underwater, as well as to perform seamless air‐to‐water and water‐to‐air transitions. A closed loop control algorithm is utilized for the vehicle's aerial and water‐surface stability and maneuver, whereas an open loop control algorithm is used for underwater maneuver. The experimental results show a fully operational prototype with six degrees of freedom underwater, stable flight, operation capabilities on water surface, and agile maneuvering underwater.     

ARDEA—An MAV with skills for future planetary missions

We introduce a prototype flying platform for planetary exploration: autonomous robot design for extraterrestrial applications (ARDEA). Communication with unmanned missions beyond Earth orbit suffers from time delay, thus a key criterion for robotic exploration is a robot's ability to perform tasks without human intervention. For autonomous operation, all computations should be done on‐board and Global Navigation Satellite System (GNSS) should not be relied on for navigation purposes. Given these objectives ARDEA is equipped with two pairs of wide‐angle stereo cameras and an inertial measurement unit (IMU) for robust visual‐inertial navigation and time‐efficient, omni‐directional 3D mapping. The four cameras cover a $240^{°}$ vertical field of view, enabling the system to operate in confined environments such as caves formed by lava tubes. The captured images are split into several pinhole cameras, which are used for simultaneously running visual odometries. The stereo output is used for simultaneous localization and mapping, 3D map generation and collision‐free motion planning. To operate the vehicle efficiently for a variety of missions, ARDEA's capabilities have been modularized into skills which can be assembled to fulfill a mission's objectives. These skills are defined generically so that they are independent of the robot configuration, making the approach suitable for different heterogeneous robotic teams. The diverse skill set also makes the micro aerial vehicle (MAV) useful for any task where autonomous exploration is needed. For example terrestrial search and rescue missions where visual navigation in GNSS‐denied indoor environments is crucial, such as partially collapsed man‐made structures like buildings or tunnels. We have demonstrated the robustness of our system in indoor and outdoor field tests.     

Experiments in unmanned aerial vehicle/unmanned ground vehicle radiation search

This paper discusses the results of a field experiment conducted at Savannah River National Laboratory to test the performance of several algorithms for the localization of radioactive materials. In this multirobot system, both an unmanned aerial vehicle, a custom hexacopter, and an unmanned ground vehicle (UGV), the ClearPath Jackal, equipped with γ‐ray spectrometers, were used to collect data from two radioactive source configurations. Both the Fourier scattering transform and the Laplacian eigenmap algorithms for source detection were tested on the collected data sets. These algorithms transform raw spectral measurements into alternate spaces to allow clustering to detect trends within the data which indicate the presence of radioactive sources. This study also presents a point source model and accompanying information‐theoretic active exploration algorithm. Field testing validated the ability of this model to fuse aerial and ground collected radiation measurements, and the exploration algorithm’s ability to select informative actions to reduce model uncertainty, allowing the UGV to locate radioactive material online.     

Active sonar-based bottom-following for unmanned underwater vehicles

The problem of high-precision bottom-following in the proximity of the seabed for open-frame unmanned underwater vehicles (UUVs) is addressed in this paper. The suggested approach consists of the integration of a guidance and control system with an active multi-hypothesis extended Kalman filter, able to estimate the motion of the vehicle with respect to the bottom profile. The guidance module is based on the definition of a suitable Lyapunov function associated with the bottom-following task, while the motion controller is a conventional autopilot, performing autoheading, autodepth, and autospeed. The motion of the vehicle is estimated from range and bearing measurements supplied by a high-frequency pencil-beam profiling sonar. Moreover, a general-purpose sensor-based guidance and control system for advanced UUVs, able to manage active sensing-based guidance and motion estimation modules, is presented. An application of the proposed architecture to execute high-precision bottom-following using Romeo, a prototype UUV, developed by the Robotics Dept. of the Istituto Automazione Navale, is described. Experimental results of tests, conducted in a high-diving pool with the vehicle equipped with a sonar profiler, are presented.     

单波束前视声纳的信息提取方法研究

介绍了一种可用于小型水下机器人的前视声纳信息提取方法。利用该方法获取了声纳视域内目标的方位信息,这在小型水下机器人的自主目标跟踪和避碰方面具有很大的应用价值。该方法主要包括3个部分的内容:利用声纳数据生成声纳图像;对声纳图像进行预处理;从处理后的图像中提取目标的特征信息。针对图像中较大目标的边缘信息,提出了一种基于最小二乘法的分段曲线拟合的方法,并给出了基于实验室的水池中获得的实测数据的拟合结果,验证了该方法的有效性。     

An Empirical Evaluation of Co-ordination Strategies for an AUV and UAV

We propose a framework for cooperative search using a combination of an unmanned aerial vehicle (UAV) and an autonomous underwater vehicle (AUV). Such a combination allows search platforms to adapt to changes in both mission objectives and environmental parameters. We propose three strategies for coordination between an UAV and AUV to maximize the area explored while minimizing the idle time of the UAV and AUV. We evaluate the efficacy of these strategies while varying the speed, communication range and the number of targets. Preliminary results suggest the feasibility of our approach to combine UAVs and AUVs for effectively searching a given area.     

基于PF和UKF的水下目标运动估计方法研究

针对无人水下航行器（UUV）目标跟踪控制需求,分别提出了水下目标的粒子滤波（PF）和无迹卡尔曼滤波（UKF）运动估计方法,建立了目标运动参考坐标系,给出了坐标系之间转换基本方法;设计了建立了目标的典型运动模型和非线性随机运动模型,利用前视声呐实测实验数据,完成水下目标运动估计。通过与扩展卡尔曼滤波器（EKF）的目标运动估计对比仿真实验,验证了PF和UKF两种目标运动估计方法的有效性。     

Underwater place recognition using forward‐looking sonar images: A topological approach

Autonomous underwater vehicles are a prominent tool for underwater exploration because they can access dangerous places avoiding the risks for the human beings. However, the autonomous navigation still a challenge due to the characteristics of the environment that decrease the performance of the sensor and the robot perception. In this context, this paper proposes a loop closure detector addressed to the simultaneous localization and mapping problem at semistructured environments using acoustic images acquired by forward‐looking sonars. The images are segmented by an adaptative approach based on the acoustic beams analysis. A pose‐invariant topological graph is build to represent the relationship between image features. The loop closure detection is achieved using a graph comparison. The approach is evaluated in a real environment at a marina. The results reveal all loop closures of the data set are detected with a high precision and present an invariant to image rotation.