DNV GL is the world's leading class company, headquartered at Høvik and with offices in over 100 countries. Here, work is in full swing on developing global technical standards for autonomous and remotely controlled ships. Challenges are lined up, but safety, environmental and economic benefits can become valuable. In this environment, we find Veronica Liverud Krathe, researcher in DNV GL's maritime research program which is closely related to technological developments.
Tomorrow's crew can work onshore
"Although we see that the term ‘autonomous ships’ has frequently been used by the media lately, evidently it will take a long time before people are completely out of the loop. Most probably, we will have a combination of autonomy and remote control, where onshore operators monitor a ship and can intervene if the ship does not handle a situation on its own. A true remote-controlled ship will need an autonomous system on board that can determine what the ship should do when losing its shore communication,” Veronica says.
“Unmanned ships remotely controlled by humans must have very reliable machinery, while fully self-propelling vessels must also have anti-collision technology, complying with the rules of the sea just like conventional ships. For example, in the absence of human supervision, image recognition will be important in order to avoid a collision, but using such technology is complicated by the fact that the rules are situational. For example, sailboats have different rules of yield depending on whether they are engine-powered or not. In such cases, machine learning and artificial intelligence can be useful,” she says.
While Norway is at the forefront of technology, with providers viewing development as a competitive advantage, we are also far ahead in developing a regulatory framework. In a global perspective, it is not the technology that is the bottleneck, but rather the regulations that take time to establish, as different parts of the world develop at different pace.
Pioneer work for tomorrow's shipping
In collaboration with among others Fjord 1 and Høglund Marine Automation, Veronica is working in a project supported by the Research Council. The Remote Operations of Machinery and Automation Systems project will look at remote operation of ship machinery from an onshore control room. By assembling machinery-related expertise and staff in such a control centre, it will be possible to consider reduced crews and skills aboard the ships. This is just one part of the approach required for remote control.
“Reliable machinery, bandwidth and data quality are prerequisites. We envision onshore control centres monitoring data from the engine rooms to a variety of ships – and other information from them – and aiding ships in need of aid,” Veronica says.
Senior researcher at DNV GL, Silje Brathagen With, believes standardized solutions are a prerequisite for achieving satisfactory data quality. As part of her role, Silje has been involved in some of the most forward-looking projects for DNV GL. One of the projects she is currently working on is to create the platform and the rules for classifying autonomous vessels. Classification is a ship certification system for safeguarding safety, quality and environmental considerations, required, for example, in relation to research in ports and harbours.
"This challenge in turn creates many exciting part projects. How to verify all the situations the ship may experience, and how to test this after the ship is built?” Silje asks rhetorically.
”For classification, this is a brand new way of working, the goal being quality assurance of complex models with an infinite number of potential outcomes.”
Silje and Veronica both emphasize three factors that drive the development of autonomous ships, requiring that rules and guidelines for such systems be implemented quickly.
It is primarily economic incentives that drive us towards autonomy.
"A large crew residing on a ship has a considerable cost in itself, so any reduction in the number of people on board is a saving. If you remove the crew completely, you can even design ships without rescue equipment or living space. This means that ships can be constructed in a more aerodynamic and hydrodynamic way," says Silje.
One of the DNV GL’s projects focuses on remote control of a ship for offshore operations, combining autonomy and remote control with humans as part of the final decision link. The Safe Implementation of Autonomous and Remote Operation of Ships project is a collaboration with Kongsberg Maritime. Veronica envisions large cost savings through such automation.
”In this project, we combine autonomy and remote control, which means that costs are significantly reduced. With sensors and solutions for information analysis on board a ship, operations can be monitored from an onshore control room,” Veronica says.
"The goal is to be safer, or as safe as today – that is what the shipping industry is concerned with. To achieve this goal, we need to figure out how to measure the safety of today's shipping,” Veronica says.
Human error occurs in all industries, but many errors are also prevented just because there are humans present. So, Veronica emphasizes the importance of determining what matters most when developing the safety of autonomous technology.
“An accident early in the development will create less confidence in the system. So, extended testing is desirable. That way we can recreate potentially dangerous situations and ensure that the ship and systems are able to cope with them," she says.
DNV GL uses its Hardware-In-the-Loop simulator to test that certified components function in the system for which they are intended.
“Developing such models is complex and demanding. They must be standardized, but one must also maintain the balance of specifications varying between projects," Veronica says.
#3 The environment
A ship may operate farther offshore without humans on board. Thus, the ship can also save fuel by reducing its speed. That will give the maritime industry a great opportunity to become more sustainable.
"When a ship is designed with only the payload in mind, and can reduce its speed, it can save significant amounts of fuel. And you will no longer have the same need for heating and lighting as on a traditional ship,” Veronica says.
“There will be no requirements to e.g. visibility and daylight. A ship without crew can therefore be delivered with more aero- and hydrodynamic design, with reduced wind and water resistance and thus reduced fuel consumption," Silje says.
Working on future solutions
Silje and Veronica find it meaningful to work on solutions with the potential to change an entire industry. At DNV GL, they will build expertise on technological solutions and frameworks for the change to take place safely, both in Norway and internationally.
"With increased focus on digitization, there is a massive change on the way throughout the industry, and DNV GL will make me change with it," Veronica says.
"Ultimately, we should not forget the small steps; the detailed solutions for the part components that all point toward autonomous ships. It's incredibly exciting to contribute to a bigger and incredibly complex whole," Silje says.
What is an autonomous ship?
An autonomous vessel has a system of automatic functions that can perform operations without any kind of human supervision.
ReVolt is DNV GL's own concept for autonomous ships. A scale model based on this concept is now being used by students at NTNU in Trondheim for testing autonomy and remote control functions.
HIL can be compared to a flight simulator for power electronics. The components to be verified are connected to a real-time digital simulator that represents the entire power system to be used. The digital simulator exposes the components to all their potential usage situations. The components' answers are then fed back.