Tuesday 28 February 2012

China - Genetic algorithms let Chinese drones hunt submarines

Chinese navy researchers have revealed how they plan to hunt submarines using ship-launched uncrewed air vehicles (UAVs).

The plan, developed by the naval academy in Dalian, China, is to choose the best hunting pattern for a drone using the power of the genetic algorithm - a search engine that evolves an optimum solution by discarding feeble offspring and breeding the best to make ever stronger ones.

The route evolved would make the best use of fuel, cater for air and sea threats and work with dropped sonar buoys. Presumably this could come in handy in some future international dispute over Taiwan. But why tell your adversary - who can now evolve counter measures?

You'd imagine that how a military hunts submarines might be a secret. In WWII, for instance, the airborne hunt for submarines with positional info gleaned from Bletchley Park's Enigma decrypts was pivotal in winning the battle of the Atlantic. But we learned about that much later. Bletchley was famously Churchill's goose that "laid the golden eggs but never cackled".

Why the department of "underwater weaponry and chemical defence" at the academy has revealed its cunning UAV plan - and published it in the journal Advanced Materials Research - is somewhat baffling.

It's not the first time this has happened. In 2010 Chinese researchers published a treatise on how to hack and trip large chunks of the US electricity grid. This led (after much initial disbelieving spluttering) to much angry rhetoric from aggrieved US commentators, not least the Department of Homeland Security.

Perhaps the frenetic, breakneck pace of China's scientific publishing machine is outstripping the nation's ability to work out what it should and should not publish?

No comments:

Post a Comment