Man bites dog: unmanned ships and the future of navigation

A few weeks before Posidonia I spoke to Roger Adamson of Futurenautics about his upcoming seminar on unmanned ships and more generally about autonomous ships for my client Global Navigation Solutions. The piece didn’t run on Great Circle in the end but I’m happy to publish it here.

In the future, so the old pilot’s joke goes, there will only be two crew in an airliner’s cockpit – a man and a dog. The man’s job will be to feed the dog, the dog’s to bite the man should he try to touch the controls. When the Economist rolled this one out last year, it was trying to put into context recent fatal crashes against an apparent trend towards greater automation in the cockpit.

Though the idea of autonomous aircraft appeals to the airlines, the newspaper noted that the casualties had appeared to spring from the nexus of man and machine at critical moments

Because pilots are spending less time flying and more time monitoring and programming equipment, automation has removed mundane chores but can cause bigger headaches when the workload is high. In the case of the two crashes, preliminary conclusions were that the pilots might have lacked the experience to deal with the situation they were in because they were used to the aircraft’s computers doing the work.

Inevitably, the potential cost savings and apparent advantages of automation have inspired the maritime industry to follow aviation into some speculation if its own. When Rolls-Royce’s Oskar Levander started mooting the autonomous ship concept last year his point seemed to be less about starting a revolution than suggesting that if we want such technology in future we should start considering it now.

The reaction – from the risk, regulation and operations community – seemed to be that he was flying a kite on a pretty long line and that the barriers to adoption were greater than the potential advantages.

Undeterred, Roger Adamson, CEO of Futurenautics hosted a seminar during the recent Posidonia exhibition considering the potential of the unmanned ship, including a contribution from GNS CEO Mike Robinson.

In publicity for the seminar, Adamson called the unmanned ships concept ‘inevitable’ but why does he believe what the rest of the industry appeared not to?

Significant benefits from autonomous vehicles, particularly in terms of efficiency and safety have already been recognised, he says and shipping needs those too. Calling the reaction to the Rolls-Royce story ‘utterly predictable PR responses from maritime’s big vested interests’ he says some large organisations are already positioning for the future, with an unmanned ship project, MUNIN, well established.

Adamson also rejects the comparison to the airline industry’s use of automation, pointing that shipping can only dream of having a safety record as good as aviation’s.

“Major incidents and loss of life in the air are tiny in comparison to the deaths, injuries and environmental damage in commercial shipping. The reality is that autopilot, or what we’d describe as automated systems, have been with us for a long time at sea.”

He accepts the criticality of human-machine interface and says that as this interaction becomes more important, ‘industrial design’ will evolve into ‘user interface expertise’ with skills in machine learning and robotics alongside fluency in communications and design.

That suggests that spending more research time and money on improving ergonomic design, system integration and training might also make bridge operations safer. That money is already being spent, but says Adamson, despite the focus on training and ergonomics, “the statistics suggest that 80-85% of all accidents at sea are as a result of human error, and that has remained stubbornly high for decades”.

The accident report on the CMA CGM Florida/Chou Shan incident last year concluded that the collision happened because VHF communications weren’t translated from Mandarin to English so the Officer of the Watch didn’t understand what was going on, one ship executed a manoeuvre which was against COLREGS and no-one knew at what point the Master should be called to the bridge. In other words a perfect storm of human errors by crews who have to fend for themselves.

“What we’re failing to really grasp is the potential of this new technology not just to automate, but to support. At the moment the Master has the sole responsibility for a huge asset, valuable cargo and multiple human lives,” he says. “The autonomous ship and the data and connectivity which are a constituent part of it will allow the ship itself to support the Master and crew. It will also allow shore-based staff, authorities, other vessels and a huge number of maritime stakeholders to have the same information the Master has at exactly the same time, and in an emergency situation, the potential to intervene if necessary.”

It’s a combination that could reduce the level of risk significantly while also spreading that huge responsibility more widely. Autonomous ship systems offer the industry a chance to attack the root causes of accidents and therefore bring down accident rates, something he says would justify a far larger investment in R&D.

The changes will happen shoreside too. With increased automation, a higher proportion of white-collar, managerial jobs are susceptible to being taken by machines in future. The concept of ‘Knowledge Automation’ driven by increasingly sophisticated algorithms with deep learning capabilities has the potential to – eventually – replace a fleet manager with an algorithm subject to minimal human oversight, he says.

“Ask yourself this question; when you have a fleet of ships which are streaming data back to shore continually and effectively sailing themselves — whether they are actually manned or not — and you can intervene from shore if you need to, how many more ships could one person manage?”

In the short term, he says autonomous ships won’t necessarily be unmanned. As more sensors are added to ships, then algorithms will be introduced to learn from that data and make decisions. Actuators will then be linked to those sensors, creating autonomous ships.

“In the same way that almost all the technology needed for a self-driving car is already in high-end cars today, commercial vessels will have the capability for autonomy well before they become unmanned. So the evolution will naturally be automation-autonomous-unmanned,” he says.

It’s an exciting prospect but one that continues to divide the industry squarely between the traditionalists and the Futurenauts. Adamson says by focussing on the benefits to safety he hopes to challenge assumptions about how close this technology is and what the potential benefits and risks are.

In that respect, the Posidonia seminar demonstrated that Adamson is far from a lone voice. Class society DNV GL chose the event to unveil its own look into “The Future of Shipping”. The report predicts both automated and remote operations relying upon the deployment of integrated systems including advanced navigation and sensor and actuator networks, and cites these as key to improving safety and efficiency.

The seminar also provoked some insight into just how sophisticated the technology already available to ship operators really is. Together with discussions about the potential impact on physical and cyber-security, the changing roles of seafarers and the crucial interaction between man and machine, Adamson’s guests and audience provided lively and thoughtful exchanges.

DNV GL believes that the first prototype of a fully autonomous ship could appear as early as 2015, with fully-automated ships entering the market by 2025 and some segments like container shipping being fully automated by 2050. “When you strip out the hyperbole and look at the reality of technology, whether we like it or not, it is difficult to argue that the autonomous or unmanned ship is about more than next-generation ship efficiency.”

A Silicon Valley start-up has technology in the blood but shipping isn’t like that. Until recently a technology focus has not been an essential part of success in the shipping and maritime industry – the lack of connectivity has played a big part in that – but that, he says, is going to change. “The technology trends we need to communicate are complex, but the unmanned or autonomous ship is a great way to show them in the context and demonstrate how relevant they are to us.”