Rolls-Royce VP of Innovation for Marine, Oskar Levander, talks to InnovOil about the AAWA initiative and the company’s efforts to put a remote-controlled and autonomous vessel to work by 2020
While it may still sound futuristic, the age of autonomous ships is already upon us. Similar applications in robotics, consumer vehicles and even mining are, in many cases, accepted practice making shipping a ripe target for disruption. As Rolls-Royce Vice President of Innovation – Marine, Oskar Levander recently remarked: “This is happening. It’s not if, it’s when.
One research group, the Advanced Autonomous Waterborne Applications Initiative (AAWA), set out its vision for the future of the sector in a recent whitepaper. Launched at the Autonomous Ship Technology Symposium 2016 in Amsterdam, this report sets out the current state of autonomous shipping technology, the opportunities it raises and the challenges that must be overcome to make it a reality.
Specifically, the group outlined the business case for autonomous applications, the safety and security implications of designing and operating remotely operated ships, the legal and regulatory dimensions and the existence and readiness of a supply chain network to deliver commercial products.
AAWA links a number of Finnish universities as well as major players in the maritime industry. In addition to companies such as DNV GL and communications specialist Inmarsat, the project is led by Rolls-Royce. At the launch, Levander commented that: “The technologies needed to make remote and autonomous ships a reality exist. The AAWA project is testing sensor arrays in a range of operating and climatic conditions in Finland and has created a simulated autonomous ship control system which allows the behaviour of the complete communication system to be explored.”
In September, InnovOil sat down with Levander to discuss some of the finer details of the AAWA report, and where Rolls-Royce itself believes this new technological leap is likely to take us.
How it works
Rolls-Royce’s plans for such vessels are clear and ambitious. “We have a roadmap for how we see these kinds of ships entering the scene. Before the decade is out we want to have the first [remote and autonomous] ship in commercial operation,” Levander affirmed. The nature of international regulations means that this first operation would likely be a smaller ship – potentially a tug or harbour boat – limited to work in the domestic waters of a nation which will grant them a flag. Given the Finnish backing for the AAWA, InnovOil believes that they may well play host. Beyond 2020, the company is keen to press on to the technological horizon. “After that we are expanding out into international shipping as soon as regulation will allow. First it will be short sea shipping and then we will go into bigger vessels and crossing oceans,” he said. From there, the future shipping landscape begins to look quite different.
Before we can arrive there however, Levander explained the basics of how these systems might work. First is his clarification that: “All these ships will be a hybrid of remote and autonomous operation. It might be more one than the other depending on application, but they will all have both.”
There are various degrees of possible vessel autonomy. As the AAWA report notes, these levels of autonomy (LOA) are described in a scale developed by Thomas Sheridan, ranging from 1 (where the computer offers no assistance and a human remains fully in charge), through 5 (a human must approve the computer’s actions) to 10 (where a computer autonomously makes all decisions with no human input).
What is important is that there are very few scenarios in which ocean-going vessels would be expected to remain at the highest LOA. Throughout its analysis, AAWA anticipates what it calls a “dynamic” or “adjustable” LOA system – as referenced by Levander – meaning all ships would likely be capable of being controlled remotely and making decisions autonomously, depending on their application.
The AAWA authors also provide a basic overview of the systems used throughout a voyage. In the voyage planning stage, operators must determine which journey legs will be undertaken autonomously, remotely or otherwise – including set failsafe or fallback strategies if any of these routes should be disrupted. Prior to embarking, the ship will also require a system to verify sea readiness – from hull integrity to securing cargo. Again, several potential LOA options would be possible when a vessel leaves port or moors, but the report indicates that most would require some level of human oversight.
Once on the water, autonomous operation becomes more straightforward. The ship proceeds along set legs and data transfer is “limited only to relevant status regarding ship’s location, heading, speed, ETA to next waypoint and key information from the ship’s situational awareness systems as well as critical ship systems,” the authors suggest. That would also enable ship operators to oversee multiple vessels at one time, providing their respective missions are proceeding as planned and further intervention is not required.
Although most situations should be navigable with a robust autonomous system – e.g. collision avoidance, evasion, or route re-planning – some may require additional intervention. In a complex navigational scenario, for example when the system must determine how to avoid a large number of other vessels, the authors anticipate that the ship will send a “pan-pan” message to an operator indicating that it requires assistance. A human user can then approve decisions or control the ship remotely. In the event that such a signal is not received by either party, the ship would be able to deploy another predetermined fallback strategy.
The nature of risk in oil and gas means that the industry is unlikely to be at the forefront in the operation of unmanned vessels. Instead, the first of these ships will be “big cargo vessels transporting non-dangerous cargo,” Levander noted, rather than new LNG carriers. But that does not mean that the oil and gas and other segments of the marine industry will not see benefits. “What we do say is that these ships will have exactly the same technology on board for improved efficiency, safety and performance,” he continued. “These tankers will opt to have the same technology but will also have a few people on board just in case.”
The routine addition of “situational awareness” technologies – networks of sensors, cameras, radars and lidar –necessary for the development of remote and autonomous ships will also improve operations and safety for those manned vessels. “Today you have radar and electronic charts and so on, but it is a very crude way of operating,” he explained. “Ships operate in fog and darkness today but there is very limited visibility. Captains rely on these kind of crude technologies, so we say that of course that an industry concerned about safety will adopt these new technologies, because it is so much better and safer than what is out there.”
Even with lower or mid-range LOA capabilities, these systems could enable hybrid scenarios, Levander added. “In certain situations you could steer a vessel, even a tanker remotely. For example, when the crew is sleeping the system can say: ‘OK we’re out in the middle of the ocean, I have as good a view here of what is happening around the ship so I can do it.’ Only if something needs attention do you wake up the crew, who remain on standby, as it were.”
As well as reducing risk, that can also help to reduce overall staffing on vessels, lowering operator costs.
Although Rolls-Royce is confident that the technology necessary to build and run these ships already exists, there are systemic improvements which could also benefit the industry. The routine use of more situational awareness technologies should improve operations overall, but the increased focus on reliability in particular – autonomous ships will need to be work more reliably than their manned counterparts if they are to be qualified – is likely to lead to much greater standardisation in an industry where few ships are identical.
There is little comparison in shipbuilding to say, aerospace or the automotive industry, where designs are made, validated and then produced. “Most ships out there today are prototypes,” he explained. “In shipping you design it, it is ordered and then it goes into use… Building these new vessels over longer periods [means that] we can afford to validate better.” That should mean vessels spend more time working and less time docked for repairs, all of which means more cost-effective operations.
In turn, smarter vessel systems also allows for better predictive maintenance. “Today we already have ‘health management’ or predictive maintenance on bigger equipment like engines and thrusters, but that is not enough for an unmanned ship. The entire system needs to be trended and reliable – so we are now going from equipment-level to system-level.” Coupled with the use of more standardised equipment, operators will be able to build a far more comprehensive image of reliability issues, and plan their maintenance accordingly.
Batten down the hatches
The glaring question with regards to autonomous and remote vessels is their susceptibility to interference or indeed cyberattack. For an industry already concerned with piracy, Levander lamented that cybersecurity is “perhaps not at the level it should be” within the marine sector. Although uncommon, the industry has seen instances where such attacks have affected vessels, even disabling bridges.
“A lot of people think that if something happens to their ship on the digital side they can always drive it manually – but that’s not always true anymore. If your bridge goes black on any vessel today, it’s not that easy to operate it manually anymore, you’re so dependent on electronic systems,” he noted.
Guarding against such attacks means building security into those systems from the ground up. “We have started incorporating this from the beginning of the entire system design and architecture, because cybersecurity is not something you can apply on top of existing things – it needs to be embedded in all parts of the operation, including the hardware side, the software side and the human aspect.”
In another example of technology transfer, Rolls-Royce brought in technologies and expertise from its aerospace division where, Levander added, thousands of jet engines have been monitored remotely and securely for years. Confidence in that technology suggests that Rolls-Royce is comfortable with the level of protection it can offer more connected vessels. Instead, Levander said that: “My concern in the future is the existing ships out there today that might not have it. They are easier targets.”
Whatever the security measures implemented, the shape and pace of future progress now largely rests with maritime regulators. With Rolls-Royce’s eyes firmly fixed on putting a remote and autonomous vessel to work by 2020, it has plenty to be getting on with. While some say it could take the International Maritime Organization (IMO) a decade to agree on governing legislation, optimists claim it could be done in four. Levander would not be drawn on a timeline beyond 2020, but said that the issuing of preliminary guidelines would offer the industry a solid platform to start from while international rules were being drafted.
What is clear is that this technology is already here, and the industry is already listening. “Since we started talking about this subject around three years ago, the interest in the marine industry has changed completely,” Levander said. “In the beginning there were a lot of sceptical voices, but during the last year everyone is talking about it. More people are sure that it is coming, more companies are working on it and more of our customers are coming to us and saying that they want it.” With any luck, they will not have to wait very long.
The AAWA Report “Remote and Autonomous Ships – The next steps” is available here.