How to get to this tomorrow is the topic of the AAWA white paper developed by Rolls-Royce in conjunction with a consortium of partners, including Finferries, ESL Shipping, Tampere University of Technology, and Brighthouse Intelligence. Part of a €6.6 million (US$7.48 million) project that runs through 2017, it identifies the areas that need to be addressed and the problems that will need to be solved to create such ships, the business case for them, and how they might be integrated into conventional shipping.
Artist's concept of a robotic ship showing a streamlined structure
One key point of the white paper is that there is no one-size-fits all solution to making an autonomous ship and nobody is going to just slap a black box on the bridge of an existing ship and hit the Go button. Such ships will appear gradually as the technology develops and, like many cloud-based systems, autonomous ships will need time to create a large database to draw from as many types of ships for many missions are built.

An element that all of these ships will have in common, according to the paper, is that they will be computer controlled with arrays of sensors that will include cameras, infrared systems, radar, lidar, microphones, sonar, and GPS. However, the tricky bit will be figuring out how to deal with all the bandwidth these require and whether they can operate in real time.
This is just one problem that would need to be solved, but AAWA says that the advantages of such ships would be potentially immense. Ships could be built without the need for crew quarters, deckhouse, lifeboats, or even decks. They would be much cheaper to build and operate and would carry more cargo. Ship design would be more flexible, human errors would be reduced, and the new technologies would provide new, disruptive business opportunities similar to those of Uber and Spotify in their fields. Central to the autonomous ships would be their ability to make decisions in what the paper calls "adjustable" or "dynamic" autonomy. That is, there are many levels of autonomy, from the lowest where the computer does nothing but follow human orders, to the highest where the computer won't even listen to humans. The ships would be programmed to select the level of autonomy suitable to the task at hand.
A remote engineer handles a propulsion problem
Most of the time, the ship is completely autonomous, such as on the high seas, and if something happens, the computer can either make its own corrections without human intervention, ask for human approval, or turn over complete control as appropriate. It's an area where the paper says the technology is already well developed thanks to autonomous cars, aviation drones, and robotics.
The paper even foresees a time when a robotic ship will be able to coordinate directly with other ships in the area. Eventually, the ships will develop the ability to learn from their own experiences and that of other ships to improve performance. But as situations become more complicated and ambiguous, even the most advanced machine will require human intervention. Weather and other conditions can change suddenly and unpredictably, equipment can break down, and the computer's programming can be overwhelmed or even deliberately attacked. In these situations, it needs the ability to revert to default and fallback positions in the event of a communications failure, with responses ranging from wait for recontact to head for a designated safe area.
The main backup for the autonomous ship would be a series of shore-based control centers linked to the ship by satellite and land-based communications. These communications would need to be bidirectional, accurate, scalable, and supported by multiple systems for redundancy and minimal risk. The operators at these stations would be able to monitor several ships at any one time, identify and correct problems in real time, act as direct contacts with human skippers as needed, and collaborate remotely on problems.

Artist's concept of a robotic container ship
Artist's concept of a robotic container ship
As to human crews, the ideal towards which Rolls-Royce is aiming is a ship with no need of one, but in the short and medium term, deckhands of flesh and blood will still play a part. At first, they may be required for legal reasons or at the insistence of insurance firms to sit about acting as emergency standby sailors or to help in dealing with conventional ships and skippers until standard practices are established. It's expected that humans will be needed in ports for longer, to ensure that cargos are properly secured.

Then there are the legal headaches of robotic ships. Is an autonomous ship legally a ship? Who is liable in the event of an accident? How must they conform to regulations or should the regulations be rewritten? What about in emergencies? Since robotic ships can't provide aid to distressed vessels directly, will it be enough for them to help with their superior sensors to guide and direct rescue vessels?
Artist's concept of a robotic cargo ship
Robotic ships may carry a token crew at first
To answer these questions and to develop the technical, legal, and safety specifications, AAWA is aiming for a proof of concept demonstrator by the end of next year. In addition, it is working on simulators, and is testing different sensor technologies onboard the FinFerries vessel Stella, which runs between Korpo and Houtskär in Finland.

"This is happening. It's not if, it's when," says Oskar Levander, Rolls-Royce, Vice President of Innovation. "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. We will see a remote controlled ship in commercial use by the end of the decade."
The video below shows Rolls-Royce's vision of the autonomous ship control station of the future.
Source: Rolls-Royce