On Tuesday, Rolls Royce and its partners in the Advanced Autonomous Waterborne Applications (AAWA) initiative released a white paper on their vision for the future of computer-controlled merchant vessels – and they believe that we will see the first such ship by the end of the decade.
"Autonomous shipping is the future of the maritime industry. As disruptive as the smart phone, the smart ship will revolutionize the landscape of ship design and operations," said Mikael Makinen, president of Rolls-Royce Marine.
AAWA's optimism for such rapid development aside, the consortium also emphasized that autonomy would proceed developmentally through phases of ever-increasing computerization of ship operations. In the area of machine intelligence, these levels are described on the Sheridan scale of human involvement in a system – ranging from full human control (as at present) to computer suggested alternatives, all the way up to full computer execution without human consultation. At first, autonomous ships may be at the lower end of the scale, giving the shipboard navigator a variety of options for him or her to choose from; later iterations could make navigation decisions and inform a shoreside operator after the fact.
The group also notes that varying levels of automation will also be appropriate for different vessel types and operational circumstances – as at present, where autopilot or autotrack navigation may be used in some areas and manual steering with a helmsman is required in others.
"In some cases, such as navigation in the open seas, the ship can be nearly fully autonomous whereas for some parts of the voyage it will require close supervision and decision making, or even full tele-operation from the human operator," AAWA wrote.
Autonomous vessels will also require a series of fallback options in the event of satellite connection interruptions or system failures. Contemplated options include: operator takes remote control; vessel navigates itself to a predesignated safe location and holding position; vessel holds position at point of system interruption. The strategy would have to be adapted to the area of navigation: "trying to maintain its position in the middle of a congested and narrow fairway in harsh weather might not be a feasible strategy," the group noted.
In some circumstances – harbors, pilotage, docking – a remote human operator would require a high bandwidth, low latency data link to monitor the vessel's position and make decisions in real time. Provisions would also have to be made for licensed pilots to take control of the vessel.
As expected, the group anticipates that a highly autonomous ship would have to be able to handle traffic effectively on its own; it would have a preset strategy for dealing with simpler situations and would notify the operator of its decisions. For more complex traffic involving multiple vessels, AAWA envisions that the ship would notify the human remote operator with an urgent message – but it would also have an automatic fallback strategy for the traffic situation, and it might even begin to execute that strategy before receiving human input.
AAWA believes that the technology to build safe, autonomous ships already exist – so what are the obstacles? First, the development costs for ship-specific, mature technical solutions; the question of how autonomy will affect liability and insurance issues; and the need to change regulations to permit the use of self-driving ships.
The group's next step will be to study these remaining questions – and to develop all required "technical, legal and safety specifications for a full scale proof of concept demonstrator by the end of 2017 and a remote controlled ship in commercial use by the end of the decade." By way of conclusion, AAWA declared that for shipping and autonomy, "the revolution has begun."