No posting

It is another national holiday today.


Ghost ships, March 8th 2014 TQ P3 (幽霊船/無人船)

Military drones already fly frequent missions and civilian operations using unmanned aircraft are coming. Driverless cars are clocking up thousands of test miles. So why not let remote-controlled ship set sail without a crew? Indeed, the maritime industry has started to think about what would be required to launch a latter-day Marie Celeste.
Ships, like aircraft and cars, are increasingly controlled by electronic systems, which makes automation easier. The bridges of some modern vessels are now more likely to contain computer screens and joysticks than engine telegraphs and a giant ship's wheel. The latest supply ships serving the offshore oil and gas industry in the North Sea, for instance, use dynamic positioning systems which collect data from satellites, gyrocompasses, and wind and motion sensors to automatically hold their position when tranferring cargo (also done by remote control) to and from platforms, even in the heaviest swells.
However, as is also the case with pilotless aircraft and driverless cars, it is not so much a technological challenge that has to be overcome before autonomous ships can set sail, but regulatory and safety concerns. As in the air and on the road, robust control systems will be needed to conform to existing regulations.
The maritime industry is interested in crewless ships for two reasons. The first is safety. Most accidents at sea are the result of human error, just as they are in cars and planes. So if human operators are replaced by sophisticated sensors and computer systems, autonomous vessels should, in theory, make shipping safer.
The second reason is, of course, cost. It is becoming increasingly difficult to sign up competent crew prepared to spend months away at sea. Moreover, some voyages are likely to get even longer for ships carrying non-urgent cargo. By some accounts, a 30% reduction in speed by a bulk carrier can save around 50% in fuel. This means slower steaming could provide big savings in fuel costs, but it would be at the expense of increased expenditure on crew for these longer voyages, both in wages and for the ゛hotel゛facilities required on board. Removing the crew, though, also removes the need for their accommodation and its associated equipment, like heating and plumbing. And that provides room to carry more cargo.
The transition to unmanned ships could take place in steps, says Oskar Levander, head of engineer and technology for the marine division of Rolls-Royce. Crews would be reduced as some functions are move onshore, such as monitoring machinery. (The engines on jet aircraft are already overseen by ground stations.) This could be followed by some watchkeeping and navigation duties. Experienced crew might be put on board when ships leave or enter port, just as pilots are to navigate. And a small maintenance crew could be kept for the voyage until remote-control systems prove themselves. A fleet of autonomous ships could also sail in convoy with a manned vessel in the lead (as illustrated on the previous page).
The onshore control rooms would keep an eye on ships thanks to live data transmitted from vessels, including video and infra-red images. Object-recognition software, combined with radar, would further automate the process. If an alarm was raised the skeleton crew on board could be alerted or the control room take charge, probably from a bridge in a simulator. Rolls-Royce already operates virtual ships' bridges, with 360 views, for the training of officers and crew. These are realistic enough to make landlubbers feel seasick.
Using onshore control rooms and simulators a team of ten land-based captains could operate 100 or so ships, reckons Mr Levander. The captains could commute from home for their shifts just as the pilots who fly military drones do. Passenger ships are likely to remain crewed, however. Trained personnel are needed to manage evacuation procedures, and in any case passengers are unlikely to want robots and vending machines attending to their needs.
The slower-sailing bulk carriers could be the first ships to be automated, according to the Maritime Unmanned Navigation through Intelligence in Networks (MUNIN) project, a European Union initiative backed by a number of industrial organisations. Like others, it says the ability of drone ships to detect other vessels and take avoiding action will be crucial, but possible with advanced technology and improved backup systems.
With a captain technically in command - even though he is based in an onshore control room - MUNIN thinks the legal and practical challenges of meeting maritime rules could met. Radio messages from other ships, along with those from coastguards and port authorities, could be automatically routed to the shore captain. Something similar is being proposed for autonomous civil drones, with ground-based pilots responding to communications and air-traffic comtrol instructions as if they were in the cockpit.
In many ways automating a ship should be a lot easier than automating aircraft, Mr Levander believes. For a start, if something did go wrong, instead of falling out of the sky a drone ship could be set by default to cut its engines and drop anchor without harming anyone. As for piracy, with no crew to be taken hostage it would be much easier for armed forces to intervene. Of course, more modern pirates might try to hack their way into the controls of an autonomous ship to take command. Which is why encrypted data communication is high on the maritime industry's list of things to do before ghostly vessels ply the trade routes.