Machine Learning News

Images: Modus Seabed Intervention

Flying new Routes Subsea

vehicle and we have developed it further so that it can be operated as a class leading survey and inspection vehicle,” says Ward. “We have focused on how the sensors are integrated, how they talk to each other, mission software and we are now moving into artificial intelligence and machine learning capabilities. We can already say ‘go here, if you see a pipeline, follow it, or if you see a monopile, perform an inspection around its circumference’. This has all been demonstrated in trials.  The next set of autonomous developments will be very exciting. In the near term

Martin McDonald, Senior Vice President, ROV Division, Oceaneering International.
Courtesy of Oceaneering International

One-on-One with Martin McDonald, SVP, ROV Division, Oceaneering

, station keeping, manipulator tasks, sensors, intervention tooling, and system diagnostics, which leads to improved performance and efficiency gains. Continued software and control systems development are key components to enable subsea residency and autonomous interventions.There’s also machine learning and machine vision. As I mentioned earlier, we have been working on automated operations, such as auto docking, where the ROV pilot can direct the ROV to move autonomously to a docking point by moving a cursor on the screen, without any intervention on the joystick. It is the machine vision

(Image courtesy of Wintershall)

4Subsea Acquires Astori

.4Subsea said it will investigate the opportunities of further digitizing Astori’s subsea control modules, in the same way it has developed digital twins for subsea production and drilling.Building on sensor data from both their own and third-party sensors and by using advanced algorithms and machine learning, 4Subsea’s domain experts provide a digital service for improved decision support to operators.“Our digital services help operators 'know it before it happens', so that they can act accordingly and perform minor repairs and maintenance, without risking replacing expensive

Marine Technology Reporter published a supplement to celebrate the 50th anniversary of Oceanology International. Photo: MTR

Oi: Tracking 50 Years of Ocean Innovation

typically built into the hull of a ship. Those instrumentation providers created a new wave of digital sonar, modern circuits, there’s been an evolution in transceivers and these shrank and have become to some extent more affordable.“The improvement in autonomy, artificial intelligence, machine learning, etc., they have matured. In the 1990s, you had to write a text file and up load it to the computer of the robot, push a button and hope it didn’t have a typo,” he adds. “Now to do a mission plan you throw a couple of way points in and say go. The software does all the hard

“Two Thales unmanned surface vessels are pictured entering Plymouth Sound for the first time.”

Photo credit: Shaun Roster Photography

Sensor and AI System for Plymouth Sound

which will provide situational awareness and communications with the unmanned vehicles operating out of the city’s harbor.The system will also provide the command & control of unmanned vehicles using the nearby ranges at sea.  Data from the network is being collected and analyzed by machine learning programs by M Subs’ growing AI team, with the assistance of Thales and a global and corporate partner leading in the field of machine learning and artificial intelligence.  The results of this work will finalize development of the computing system at the heart of the maritime AI system

Zooglider (top) with a selection of zooplankton imagery the robot has captured. Top photo: Benjamin Whitmore

New Robot Can Sense Plankton Optically and Acoustically

to make Zooglider,” Davis said. “Even more gratifying was to see how new facts of zooplankton behavior were deduced from its data."Co-author Jeffrey Ellen from UC San Diego’s Jacobs School of Engineering is creating methods by which Zooglider image data can be analyzed through machine learning. Other contributors include Kyle Grindley, Benjamin Whitmore, and Catherine Nickels of Scripps.Development of Zooglider was supported by the Gordon and Betty Moore Foundation

(Image: Vissim)

Equinor and Vissim Partner

Vissim said it has reached a global frame agreement with Norwegian energy giant Equinor to deploy its solution for surveillance of the ocean space surrounding all Equinor- and several third-party-operated assets on the U.K. and Norwegian Continental Shelves. The new system will incorporate machine learning and enable surveillance of vessel traffic (VTMS), air traffic, subsea structures, environment and real time weather situation and forecasts.Conceived to incorporate both traditional oil and gas and offshore wind assets, the resulting system is one of the largest of its kind in the world, Vissim

(Photo: Rovco)

Rovco Awarded Funding for 3D Visualization Project

continues from the Phase 1 project and trials proving the technology at the Offshore Renewable Energy (ORE) Catapult’s test center. In collaboration with the ORE Catapult test center and ScottishPower Renewables, Rovco will conduct additional research into the 3D vision technology by applying machine learning and AI based analysis to 3D survey and cable burial data. During the 18-month project, Rovco aims to achieve live transmission of 3D subsea models to shore.Additionally, as the project progresses, Rovco will commercialize its technology.Brian Allen, Rovco chief executive and founder, said, &ldquo

CARIS Onboard can be used to process and monitor survey data from multiple platforms (Image: Teledyne CARIS)

Teledyne CARIS to Present at Hydro 18

the volumes of data generated by modern survey operations, ensure it is of suitable quality, and turn it into more information products to support our oceans. Space is limited in this information session, so register early to ensure your seat.‘Advancing Hydrographic Data Processing through Machine Learning’ on October 30, 5:30 p.m. presented by Burns Foster, Manager of New Product Initiatives, Teledyne CARIS. Attendees will learn how this newly assembled machine-learning algorithm is being used to automatically identify and classify common types of noise in sonar systems to significantly

SeaScout Expeditionary Seabed Mapping and Intelligence System deployed during ANTX2018 (Photo: Kraken Robotics Inc.)

Kraken, ThayerMahan Demonstrate SeaScout for US Navy

test and evaluation carried out since both companies signed a strategic alliance agreement in February 2018.U.S. Navy Vice Admiral Michael Connor (retired), ThayerMahan’s President and CEO, said, “It’s our fundamental mission to integrate and deploy the latest sensors, robotics and machine learning technology as affordable and innovative solutions to the United States and allied navies. I’m very pleased that during ANTX ThayerMahan and Kraken demonstrated superior resolution and range relative to competing systems costing many times more; live streaming of target detection results

(Photo: Olis Robotics)

Olis Robotics Acquires Machine Learning Startup

Olis Robotics, developer of next-generation remote robotics software, announced it has acquired Seattle-based machine learning startup White Marsh Forests (WMF).Olis, formerly BluHaptics, has created a next-generation software platform built from the ground up that expands the capabilities of pilot-controlled service robots in the energy and space sectors and other harsh operating environments. The firm's technology enables vast improvements in robotic dexterity, precision, efficiency and overall mission success.The addition of machine learning capabilities from WMF allows robots to recognize and

(Image: Olis)

New Remote Robotics Manipulator Control System Launched

technology company, offering intuitive controls and more efficient operations which combine to reduce risk, increase precision and save time.Don Pickering, CEO of Olis explained, “This product is just the beginning of how Olis is redefining the future of remote robotics. By introducing machine learning and variable autonomy in the very near-future, we will unlock broad and diverse new marine use cases. The ES1 Control System will make it safer for pilots to perform intervention tasks in extreme and dynamic environments by making the robots they control smarter.”Key ES1 features eclipse

MTR does not present an “MTR100 Creative Photo” award, but if we did this year’s winner is Houston Mechatronics. Pictured is Houston Mechatronic’s Aquanaut in wet testing earlier this year holding it’s MTR100 ‘trophy’. (Photo: Houston Mechatronics)

MTR100: The Ones to Watch

Based in Aberdeen, Elaine Maslin is MTR’s ‘eyes & ears’ on the ground scouting new tech for our pages. Her mandate for the MTR100 was simple: ‘find five companies to watch closely in 2019.’ This is what she found.Forssea: Leveraging Machine Learning & AIMachine learning and artificial intelligence have made a huge impact on our day to day lives, from image recognition in Facebook to cars parking for themselves. These technologies are entering the subsea space in various shapes and forms.Paris-based Forssea Robotics is a subsea robotics and research and

TCarta will deploy machine learning and computer vision techniques to enhance satellite derived bathymetry in the littoral zone. (Image source: Copernicus Sentinel data 2018)

NSF Grant to Enhance Satellite-derived Bathymetry Technology

grant by the National Science Foundation (NSF) to enhance and automate multiple techniques for deriving seafloor depth measurements from optical satellite imagery. The ‘Project Trident’ research seeks to transform existing satellite derived bathymetry (SDB) techniques by leveraging machine learning and computer vision technology to enable accurate depth retrieval in variable water conditions.If successful, TCarta said, these enhanced bathymetric techniques will have positive impacts on operations related to oil and gas exploration and production, coastal infrastructure engineering, environmenta

Figure 1: The first iteration of the Tentacle is designed for Kraken’s KATFISH. (Image: Kraken)

Meet Kraken’s Tentacle Autonomous LARS

effort to mimic the memory of a typical human, the onboard memory of the Tentacle Winch records all motion data within the system and any significant motion events (over temperature, excess torque, excess tension, etc.) This data can be used to automatically tune the control system modules through machine learning techniques, improving the response times and overall performance, furthering conditioning the reflexes of the Tentacle.One example future application under consideration is supply ships and other vessels that experience relative motion issues. These vessels would benefit greatly if the motion

(Photo: EOMAP)

Looking to the Future of Satellite-derived Bathymetry

alongside with traditional survey methods,” said Dr. Marco Filippone, Chief Hydrographer at Fugro. “We can use SDB to augment existing technology as a benchmark for high definition data sets and with this new technique develop together enabling technology, processing workflows and machine learning – and we can really speed up the process providing the final users with a product that can be used for their needs.”The next SDB Day will be announced shortly

Terry Makewell  (Photo: UKHO)

UK Hydrographic Office Appoints Makewell CTO

and gained an MBA (Entrepreneurship) in 2006.   One of Makewell's first engagements at the UKHO was to support its data science experts who presented at the Government Digital Service’s ‘Sprint 18’ conference on May 10th. Here they demonstrated how the UKHO is using machine learning to exploit satellite imagery in order to help protect the marine environment and promote global maritime trade.  

Marine Technology Magazine Cover Apr 2019 - Ocean Energy: Oil, Wind & Tidal

Marine Technology Reporter is the world's largest audited subsea industry publication serving the offshore energy, subsea defense and scientific communities.

Subscribe
Marine Technology ENews subscription

Marine Technology ENews is the subsea industry's largest circulation and most authoritative ENews Service, delivered to your Email three times per week

Subscribe for MTR E-news