This paper presents a novel approach for estimating the relative motion between two moving offshore vessels. The method is based on a sensor fusion algorithm including a vision system and two motion reference units (MRUs). The vision system makes use of the open-source computer vision library OpenCV and a cube with Aruco markers placed onto each of the cube sides. The Extended Quaternion Kalman Filter (EQKF) is used for bad pose rejection for the vision system. The presented sensor fusion algorithm is based on the Indirect Feedforward Kalman Filter for error estimation. The system is self-calibrating in the sense that the Aruco cube can be placed in an arbitrary location on the secondary vessel. Experimental 6-DOF results demonstrate the accuracy and efficiency of the proposed sensor fusion method compared with the internal joint sensors of two Stewart platforms and the industrial robot. The standard deviation error was found to be 31mm or better when the Arcuo cube was placed at three different locations.

/pdf/relative-vessel-motion-tracking-using-sensor-fusion-aruco-1ncfyd9uem.pdf

Relative Vessel Motion Tracking using Sensor Fusion, Aruco Markers, and MRU Sensors

Radars with mmWave frequency modulated continuous wave (FMCW) technology accurately estimate the range and velocity of targets in their field of view (FoV). The targeted angle of arrival (AoA) estimation can be improved by increasing receiving antennas or by using multiple-input multiple-output (MIMO). However, obtaining target features such as target type remains challenging. In this paper, we present a novel target classification method based on machine learning and features extracted from a range fast Fourier transform (FFT) profile by using mmWave FMCW radars operating in the frequency range of 77–81 GHz. The measurements are carried out in a variety of realistic situations, including pedestrian, automotive, and unmanned aerial vehicle (UAV) (also known as drone). Peak, width, area, variance, and range are collected from range FFT profile peaks and fed into a machine learning model. In order to evaluate the performance, various light weight classification machine learning models such as logistic regression, Naive Bayes, support vector machine (SVM), and lightweight gradient boosting machine (GBM) are used. We demonstrate our findings by using outdoor measurements and achieve a classification accuracy of 95.6% by using LightGBM. The proposed method will be extremely useful in a wide range of applications, including cost-effective and dependable ground station traffic management and control systems for autonomous operations, and advanced driver-assistance systems (ADAS). The presented classification technique extends the potential of mmWave FMCW radar beyond the detection of range, velocity, and AoA to classification. mmWave FMCW radars will be more robust in computer vision, visual perception, and fully autonomous ground control and traffic management cyber-physical systems as a result of the added new feature.

Classification of Targets Using Statistical Features from Range FFT of mmWave FMCW Radars

An increased level of complex offshore load handling operations is expected due to an increased amount of floating wind turbines, remote fish farms, and autonomous shipping, and in general more advanced operations to be carried out at sea. A common problem for these applications is that both equipment and personnel have to be transported between two floating vessels at sea. An investigation of the Vessel-to-Vessel Motion Compensation (VVMC) problem may increase the efficiency and safety of such operations in the future. In this paper, a control algorithm has been developed and experimentally tested in the Norwegian Motion Laboratory featuring two Stewart platforms (SPs), an industrial robot, two Motion Reference Units (MRUs), and a Maritime Broadband Radio (MBR) used to establish a real-time wireless communication link. The resulting overall compensation error is measured to be no more than 41.9 mm during the experiments.

Inverse kinematic control of an industrial robot used in Vessel-to-Vessel Motion Compensation

Resilient position, navigation, and timing (PNT) data is indispensable information in the field of e-navigation. An automatic identification system (AIS) based ranging mode (R-Mode) is put forward to develop a terrestrial backup system in order to overcome the vulnerability of the global navigation satellite system (GNSS). In general, at least three base stations are required in the traditional R-Mode positioning method. However, the geometric distribution of existing base stations is not considered for positioning, as AIS is a communication system. In some cases, a vessel can only receive signals from two base stations. In this paper, an improved position estimation method based on displacement correction is therefore proposed to solve this problem. Compared with the prior displacement correction position estimation (DCPE) method, the proposed method can improve positioning accuracy effectively by adopting a more precise motion model for the vessel, including an accelerated motion and a turning motion model. Moreover, the motion model is employed adaptively to correct the displacement of the vessel. Finally, the proposed method is verified and the performance is analyzed and compared by simulation. This study can extend the application region of AIS R-Mode.

An Improved Positioning Method for Two Base Stations in AIS

The industry state-of-the-art offshore motion compensation is today carried out by offshore cranes equipped with Active Heave Compensation (AHC) systems, meaning that a hanging load is positioned at a certain height above the seafloor, even though the floating vessel carrying both the crane and the load is moving due to wave disturbances. Lately, there has been an increased interest among the industry also to compensate for the movement of a secondary vessel, which indicates that a more detailed investigation of the Vessel-to-Vessel Motion Compensation (VVMC) problem is required. The primary motivation for such compensation technique is influenced by the demand for load transfer between ships during harsh weather conditions, service on floating wind turbines and fish farms, and the transportation of personnel and equipment safely between two floating vessels, for instance. In this paper, the capability of the Leica AT960 laser tracker to measure the relative position and orientation between the two moving Stewart platforms simulating the vessels movement at sea is investigated. The acquired measurements are used to carry out a real-time VVMC using the experimental lab setup in the Norwegian Motion Laboratory at the University of Agder. The obtained experimental results and motion compensation effectiveness are quantified in terms of the accuracy and real-time performance.

Real-time 6-DOF Vessel-to-Vessel Motion Compensation using laser tracker

Relative Vessel-to-Vessel Motion Compensation (VVMC) between two floating vessels can be considered an extension of existing Active Heave Compensation (AHC), which is current state-of-the-art in the offshore industry. In this paper, the experimental results from a small-scale experimental setup of a VVMC scenario are presented. The work focuses on the kinematics related to VVMC, sensor accuracy of Motion Reference Units (MRUs), and the communication delay related to wireless transmission of MRU data between two floating vessels. The accuracies of the MRU sensors are good and verified to be less than 5% of the heave amplitude. The wireless communication link had a typical time-delay of 8ms. The presented work demonstrates the 3D-kinematics related to VVMC using homogeneous transformations, and the experimental results show that wireless transmission of MRU data is a viable solution for offshore VVMC scenarios.

Testing of wireless sensor performance in Vessel-to-Vessel Motion Compensation

Bollard is a vital component of mooring system. It is the anchor point for mooring ropes to be fixed in order to secure the vessel or ship. An algorithm that translates the segmented mask of bollard output from masked R-CNN along with bounding box and associated class probability to its corresponding edge coordinate and finally to the single reference point for efficient detection and classification of bollard towards autonomous mooring is presented. At first stage, Mask R-CNN framework is trained with custom built bollard. The model obtained from the training is inferred with real data resulting in instance segment of bollard. The segmented mask obtained contains relatively large amount of the data points representing the whole area of bollard, which typically is not desirable. In order to precisely localize the bollard with one reference co-ordinate, the proposed algorithm is applied to segmented mask. Firstly, it translates the segmented mask to only four co-ordinate points, where each point correspond to the edge of bollard. Further, from the edges, the reference point is estimated. This causes significant reduction in point of interest (POI) and has potential to reduce the error encountered during pose estimation of the bollard in 3D thus making the autonomous mooring more precise and accurate.

Per-Ove Lovsland

Papers

Testing of wireless sensor performance in Vessel-to-Vessel Motion Compensation

Autonomous Mooring towards Autonomous Maritime Navigation and Offshore Operations