Robot ships and unmanned autonomous boats
Image credit: Rolls Royce
We have robot planes, trains, cars, tanks, small boats, and now, it seems, it’s time for autonomous ships.
China has unmanned naval boats patrolling the South China Sea. Royal Navy Commander Peter Pipkin says that autonomous boats will become as important as unmanned aerial vehicles in military operations, particularly when it comes to minesweeping and anti-terrorist patrols. Researchers use small autonomous boats to conduct scientific research, and oil and gas companies use them to conduct surveys and clean up spills. Now it seems that much larger autonomous ships are on the way. The US has begun testing its first unmanned warship, and in June Rolls-Royce spoke of plans for an autonomous remote controlled cargo ship.
Rolls-Royce says the ship will be ready for 2020. That’s the 400th anniversary of the Mayflower’s voyage from Plymouth to the New World. Plymouth University and Shuttleworth Design are currently working on a solar-powered remote-controlled ship that they plan to sail across the Atlantic to commemorate this voyage.
Jay McFadden, Rolls-Royce senior vice president, marine services - engineering and product management, explains that unlike the smaller military ships, which are deployed from other manned vessels and only interface with the launch and recovery system, the larger autonomous cargo ship will have the capability to integrate autonomously with the rest of the infrastructure - to moor and dock in port, unload and offload cargo.
Richard Daltry, technical director at ASV Global, which designs smaller autonomous surface vehicles, adds that the technology under consideration is basically the same as that used in self-driving cars, but optimised for use at sea.
At the moment, the proposed Rolls-Royce ship looks like a cross between a gigantic torpedo and a metal whale. It has about as much in common with modern-day cargo ships as today’s ships have with the Mayflower.
New take on an old concept
The actual idea of a ship without crew is not new, though. As far back as 1973, Rolf Schonknecht, in his book ‘Ships and Shipping of Tomorrow’, envisaged a future where a ship’s captain performed his duties in an office building onshore, while an on-board computer navigated the ship.
In fact, to some extent, the idea of crewless ships is the culmination of a process that has been going on ever since humans put the first boat in the water. As technology has advanced, ship builders have gradually replaced crew members with mechanical parts.
During the 1980s, the idea of intelligent ships operating without crews was discussed in Japan. The idea was discarded, however, when cheaper foreign crews became available to Japanese shipping companies.
In the 1990s, renowned ship designer Kai Levander, father of Rolls-Royce vice president of marine innovation Oskar Levander, argued that a ship could travel short distances without a crew, aided by GPS and guided from traffic stations. Only when the ship neared port would a pilot board to complete the last leg of the journey.
Around the same time, German naval architect Volker Bertram suggested that a combination of artificial intelligence (AI)and teleoperation was feasible for ships. Back then, however, the concept was still not attractive to shipping companies due to high maintenance costs.
The concept re-emerged in a 2007 paper on the future development of the maritime industry by Waterborne TP, a cluster of European maritime stakeholders. Although this paper suggested that more advanced automation and improved sensors might be desirable, it stopped short of advocating full automation.
Five years later, inspired by this idea, European research groups launched the collaborative MUNIN project, co-funded under the EU’s FP7 research programme.
MUNIN stands for Maritime Unmanned Navigation through Intelligence in Networks. In Norse mythology, Munin is the name of a raven belonging to the god Odin, which each day flies around the world without any guidance, gathering information, and in the evening safely returns the information, its ‘cargo’.
For three years, European experts looked into the possibility of developing a viable concept for an unmanned and autonomous merchant vessel. Areas covered included autonomous navigation, engine and propulsion automation, communications and shore connectivity, system redundancy and efficiency, as well as questions related to regulations and liability.
Project MUNIN concluded that such a vessel was possible, but only for deep-sea voyage and not in congested or restricted waters, where a crew should operate the ship. Restricted satellite bandwidth in certain regions and high communication costs, project leaders said, makes a remote-control solution unattractive. They decided that the new ship’s monitoring and controlling functionalities would be executed by an operator in the shore control centre.
Things are still at the research and development stage, but Rolls-Royce is convinced that it will have a fully autonomous ship in the water by the end of the decade: a ship controlled via computer from a land-based hub. “The technology to do this now exists,” Oskar Levander says.
Benefits for shipping companies
Rolls-Royce argues that with no crew and therefore no crew costs, the boat will be cheaper to run. Without the infrastructure needed to support human life, such as air conditioning, heating, sanitation and crew quarters, there will be more room for cargo. The company claims that such a boat would be 5 per cent lighter and use 12 to 15 per cent less fuel.
Rolls-Royce has also created a simulation video showing how the control system might work. In the film, a small crew monitors and controls the operation of a fleet of vessels around the world. The crew uses interactive smart screens, voice recognition systems, holograms and surveillance drones to monitor what is happening both on board and around each ship. It looks like something devised for a superhero film, although McFadden says that Rolls-Royce already has most of the technology it needs. That’s the communication systems, a series of networked simulators, which is essentially the same technology that would be used in remote operations of the equipment, and a control system, which enables the gathering, processing and communication of the on-board information to a remote centre or into a computer that will create the autonomous actions needed to run the vessel.
“The challenge ahead is integrating all this into a cohesive system that provides the safety and reliability that’s necessary for remote and, eventually, autonomous vessels,” McFadden says.
Richard Daltry, technical director at Autonomous Surface Vehicles, adds that navigation and control technology is transferable from small autonomous boats, which are relatively common, to bigger ships. “The rules that apply to the vessels are similar, whatever the vessel size.”
Daltry adds that the biggest difference is safety. “The risk to other people and the environment presented by an oil tanker is significantly greater than by a 30ft [9m] unmanned oil and gas support vessel,” he says. “The level of scrutiny and regulation imposed on the software and electronics is likely to be significantly higher.”
Other companies are in the process of designing unmanned and autonomous ships. Robert Allan Ltd has a plan for a remote-control tugboat, and DN-VGL is looking at a 60-metre battery-powered autonomous ship designed for short journeys. In August 2015, China announced that it too is conducting research into autonomous ships.
All this sounds very exciting, but not everyone in the marine industry is convinced.
“It might be good from an owner and manager’s point of view in that it costs less,” says George Wilson, a marine engineer at Charles Taylor Adjusting in London. “But what happens if it goes wrong? I don’t believe that this has been heavily explored and allowed for in any of the systems that we have to date.”
Ian Dunbar, a marine engineer from Meridian Maritime Consultancy, Newcastle, adds: “I’m in favour of any ship running more efficiently, but not to the detriment to the personnel on board and the environment, should something go wrong.”
Marine engineer John Gouch from BCTQ Maritime Consultants thinks it’s likely that things will move towards automation in the coming years. “There will always be situations where you need some human intervention, though,” he says, “whether that’s remote in the same way as people use drones or whether they look at transferring a crew, when needed. Harbour masters will be concerned about the risks of coming into port and as a deep-sea engineer I know that at sea, you’re dealing with minor and major breakdowns almost daily. To deal with this, there has got to be an absolute redundancy or high-quality systems.”
Gouch adds that with current technology, we’re not a million miles away from getting a ship with those qualities.
Ships have to be able to deal with bad weather, pirates and - with so much computerised technology on-board - hackers. McFadden believes each of those things provides its own unique challenge. “Take pirates,” he says. “The target is actually attractive to pirates because it is manned. The pirates can hold hostages for ransom. An unmanned vessel would be a lower priority target. If a boat was to be attacked by pirates, we could shut off all the power and propulsion, make it harder for pirates to bring the vessel and its cargo ashore. We can alert the authorities and install homing technology so the ship automatically travels to a predesignated safe port.”
Daltry adds that ship designers will try and mitigate against things that humans solve, by having back-up solutions that are actually part of the system. “You can have greater pump capacity to deal with flooding and double hulls in place to reduce impact,” he says. “With no people on-board, you can use more aggressive firefighting means, shut down oxygen supplies. You’d also have a means to bring people on-board if it’s safe to do so, say a helicopter landing pad.”
Wilson isn’t convinced. “Having seen some of the things that can go on at sea, I’m not sure how everything can be adequately covered by a monitoring system,” he says. “There are modifications to fire alarms now, two different types of detector head. The system will only start its firefighting operations if both these go off, but could you do the same with a bilge? What happens if a pump seal fails and suddenly the bilge fills up and starts pumping water, the pump seal fails, the engine room fills up, the bilge tank fills up? You can’t pump it over the side. What happens then?”
Gouch doesn’t think the concept of autonomous ships is fully cracked at the moment. “People will probably tolerate it across oceans and clear waterways, outside of a port environment,” he says. “There’s AIS [automatic identification system] and radar, enough to prevent collisions, and the vast majority of collisions are caused by human error anyway. Over time, people might then get more used to autonomous vessels coming into busy shipping areas.”
Dunbar adds: “There’s going to have to be a lot of thought and a lot of proving to the relevant authorities before anything will happen.”
Existing maritime regulations were conceived long ago, when no one had ever thought of, let alone started discussing, autonomous ships. There are a lot of regulatory bodies covering everything from ship design to jurisdiction. To get something this new and different through all those committees could take a very long time. That’s without persuading insurance companies to insure the crewless ship and its cargo.
McFadden admits there are still issues to solve. “We don’t yet have a perfect way of handling every circumstance,” he says.
Rolls-Royce is testing the sensor arrays that it hopes will navigate its autonomous ships. McFadden adds that the company is also considering the use of drones to inspect ships more quickly, after Airbus demonstrated that this could work with aircraft at the 2016 Farnborough Air Show.
It’s important to remember, though, that this is still in the research and development stage. No one has an autonomous ship ready to make its maiden voyage yet.
With so much at stake should something go wrong, no one is likely to risk doing that either. Not until they’re sure that it’s safe and that it’s going to work.
In 1998, the US Navy also claimed that its new Yorktown smart ship would save manpower and maintenance costs. However, on its first voyage, the ship ground to a halt after bad data was fed into its computer causing the propulsion systems to fail. The navy had to tow the ship to port where engineers spent two days fixing the problem.
What was that about technology only being as good as the people using it? *
Ships without crew
The MUNIN research project concluded that an autonomous ship would need:
- An advanced sensor module for lookout duties on-board the vessel,continuously fusing sensor data from existing navigational systems. This could be radar and AIS (automatic ID system), combined with daylight and infrared cameras.
- An autonomous navigation system that follows a predefined voyage plan, but with a certain degree of freedom to adjust the route in accordance with legislation and good seamanship.
- An autonomous engine and monitoring control system, which enriches ship engine automation systems with certain failure-pre-detection functionalities.
- A shore control centre that continuously monitors and controls the vessel after it is released from its crew.
- A shore control centre operator who monitors the operation of several autonomous ships at the same time from a desktop cubicle station and controls the vessels.
- A shore control centre engineer who assists the operator in case of technical questions and who is in charge of the maintenance plan for the vessels.
- A shore control centre situation room team that can take over direct remote control of one vessel in certain situations via a shore side replica of the unmanned vessel’s bridge.