The prediction of marine fuel consumption and ship exhaust gas emissions are indispensable to evaluating ship sustainable performance under current shipping fuel standards. Big data with evolved machine learning techniques have been proved to be an effective way to contain uncertainties for ship activities. This work collects the latest global LNG carrier fleet with 435 data points and attempts to predict the marine fuel consumptions and ship-resulted global warming potential (GWP) gas emissions, including CO 2 , CH 4 , N 2 O, and black carbon aerosols. Gaussian process regression and ensemble machine learning approaches, to achieve this goal, are employed to infer the relationship between predictors (i.e., dimensional parameters, machinery parameters, and tonnage) and response variables (fuel consumptions and GWP exhaust gas emissions), providing exceptional insight into ship sustainable solutions. To improve the prediction accuracy, the hyperparameter optimization analysis via random search and Bayesian optimization is adopted to find the optimal machine learning model. The appealing results are in line with the validation data, illustrating high effectiveness and robustness of the proposed machine learning models. The procedure established in this study presents a novel approach for accelerating the research and development of sustainable shipping fuels under normal ship activities.

In this work, we attempt to illustrate possible applications of automated object detection in video sequences and still images using state-of-the-art artificial intelligence methods, such as convolutional neural networks. These novel tools can be used in various application domains in offshore operations. Neural networks can be trained to detect humans, specific objects and even learn how to conduct pipeline, anchor, tank and other types of inspections where they can significantly reduce the possibility of human error, not by replacing the human operator, but by helping him or her to be more effective and efficient. One of the problems we will discuss is how to harness the power of image pre-processing methods to increase the informative content of inputs, especially in environments where video or still-image visibility may be poor due to weather conditions, underwater turbidity or smoke.

In response to the lower for longer oil price environment and the rapidly changing digital opportunities, agile development and deployment of digital solutions using data analytics can become essential in collecting short term business value through cost reduction and building flexible technology infrastructure for long term digital transformation. The new agile development/deployment adopted by the company consists of the following process: identify operational need, develop digital solutions through a combination of conventional physics based model and machine learning techniques, deliver through a flexible real-time web platform, and feedback learnings from deployment to facilitate next round of development. ReamerVision, the real-time early detection of under-reamer failure during hole enlargement while drilling at deepwater Gulf of Mexico is one of the successful examples: – Operational need identified: Two active cutting structures (drill bit and expandable under-reamer) are often used for hole enlargement while drilling in Deepwater. A potential operational failure is the under-reamer may subject to premature cutter damage while drill bit is still in good condition. Without detection of this failure, an undergauge hole might be drilled and causes difficulty for subsequent casing running. – Data analytics methodology: Historical data of both surface and downhole mechanics data are collected. Several metrics based on physics model are constructed to quantify cutter conditions, drilling efficiency, and formation variation. Supervised learning of historical dataset processed by those metrics yields thresholds for failure detection in real-time. – Deliverable format: a web-based real-time data platform is constructed for deployment of this technology. A workflow is established between remote control center, drilling engineers, and domain-experts to establish real-time monitoring of under-reamer condition and provide actionable information to operation. – Continue improvement: since April 2016, ReamerVision has been deployed to 27 deepwater reamer runs as a risk-mitigation tool. Positive feedbacks from end-users are utilized for calibration of model and development of new features.

Water turbidity is a frequent impediment for achieving satisfactory imaging clarity in underwater video and inhibits the extraction of information concerning the condition of submerged structures. Ports, rivers, lakes and inland waterways are notoriously difficult spots for camera inspections due to poor visibility. This problem motivated us to study methods to extract a cleaner image /video from the one acquired in an almost real-time setting (delay of the order of 6–7 secs). This type of problem arises in image post-processing as an illumination neutralization problem and, it can also be viewed as a blind deconvolution problem. We present a method which enables the derivation of a cleaner image from a poor visibility original by means of a combination of linear and non-linear deterministic mathematical transformations for illumination neutralization implementable in almost real-time on GPUs. Real time visibility improvement for marine and water environments is a suite of algorithms aiming to restore the visibility of images and videos acquired under the surface of every water body utilizing our illumination neutralization method. We are currently transitioning from an academic algorithmic suite to a product we will call αλσvision (ALSvsion) from the Homerian Greek word αλσ which means sea. Upon completion of the project, the GUI will enable the view of the original camera feed and of the visibility improved video side-by-side. This software is not intended to replace the original video feed, but to offer guidance for interpreting it. αλσvision can be used for a variety of visual inspections in marine and offshore industries. The method we present also works for visibility restoration in still images or videos acquired in poor atmospheric conditions such as fog, haze, smoke or with insufficient illumination. One of our main contributions is the development of a mathematical theory which enables the derivation of results showing that lines and textures significant for the identification of structures and of structural problems are made visible as they appear in an image or video acquired in good imaging conditions.

A finite depth 3D Green function has been developed to estimate wave loads on the floating body in frequency domain which is then coupled with a time domain motion simulation tool. Nonlinear hydrostatic and Froude-Krylov forces have been incorporated considering the instantaneous wetted surface and the nonlinear equation of motion is then solved considering the large-amplitude rotations of the body. However, the option of using a consistent second order forcing model is also retained in the code where the linear diffraction forces and moments are augmented with the drift forces calculated from Newman approximation of difference frequency quadratic transfer functions (QTF). The main advantage of the method lies in its capability to simulate large-amplitude motions of ship shaped and non-ship shaped vessels with or without mooring and with or without forward speed in deep or shallow water depths. This allows investigation of various nonlinear effects and ensures safe operation. It is particularly useful in modelling the motions of a ship in a port channel or towing of offshore platforms from fabrication yard to deep water installation site. In this article, a methodical approach in validating the numerical results against published and other established numerical programs is presented for the developed finite depth Green function and hydrodynamic coefficient predictions. Also the drift forces and moments obtained by different forms of the Newman approximation are compared and validated against commercially available software. The ability of the time domain simulation program to capture nonlinearities is also shown by simulating the Mathieu type instability in heave and pitch modes of a Single Column Floater.

The underwater welding is the emerging technology of the present and the future. Many offshore platforms and ships need the underwater welding process during emergency. There exist many research institutes globally in developing techniques for underwater welding. A dire need for developing the underwater vehicle for welding processes in marine applications. The recent developments in the underwater vehicles have led to a revolutionary change in the design and system development for coastal surveillance, sea bed inspection and ocean bed exploration. This paper presents a review in implementation of a robot arm for an autonomous underwater welding vehicle. It deals with review of various underwater welding techniques for marine applications and structured such a manner all aspects of underwater welding process with robot manipulator as per the classification requirements. A brief review of its characteristics, applications and the risk mitigation factors of underwater welding techniques for marine applications. The end result is a model that provides accurate predictions for the joint torques of a two-link robot arm in a simple form implemented with algorithms for precision planning and control during underwater welding application.

This paper describes the design of marine current turbine blades based on the results of a fully-aligned lifting line optimization model. The geometry is generated by assuming chord and camber distributions in order to reproduce a given loading on the turbine blade. This geometry is then analyzed with two three-dimensional potential flow solvers: a low-order panel code and a vortex lattice model. Accuracy and stability for different discretization schemes is studied, as well as the effect of different wake alignment strategies.

In this paper, the performance of a horizontal axis marine current turbine is predicted by three numerical methods. The predictions are compared with the experimental measurements for the same turbine model. A fully unsteady wake alignment is utilized in one of the methods in order to model the realistic wake geometry of the turbine. A database-searching design method is proposed to improve the turbine performance by coupling a nonlinear optimization method with one of the numerical analysis methods.