The Inquirer-Home

Euroboffins developing trustworthy Skynet

Reliability is the key
Thu Jul 08 2010, 15:14

CATASTROPHIC FAILURE seems to be the utilities industry bogeyman, and rightly so. Running a continent-wide network of power plants and keeping them all running while you proceed to upgrade critical systems - and all while you're praying some software bug won't turn your world upside-down - is an ambitious proposal.

Enter the European Union. Through some carefully wielded funding, IBM Israel, ABB Switzerland, VTT in Finland and IAI in Israel have joined forces with universities throughout Europe to weather the impending bug apocalypse. The research project, dubbed PINCETTE, meaning "tweezer" in French, relies on the collective brain power of universities in the UK at Oxford, Turkey, Finland, Italy and Switzerland, and its sole purpose is to develop tools to improve networked systems' reliability in critical operations, work out bugs and get things up and running in next to no time.

It seems upgrading and certifying new systems while integrating them into existing networks is a huge and understandable headache that is likely to run bills into the millions of euros. This, on the other hand, will make sure that nothing goes awry, toppling over computer after computer on the network and crashing your entire fleet of Eurostars. According to IBM, the three-year project will help develop a networked system that will keep itself in check while old hardware and software elements are upgraded, guaranteeing backwards compatibility and new features alike.

The idea is that a certification system will be engineered that will allow trustworthy upgrades time on wide-scale operations such as nuclear plants or rail transportation without any breaks in the system. IBM says that this will save lots of money that is currently lost to downtime and debugging.

If they get it right, it sounds to us like a good foundation for a self-aware network of computers. µ

Share this:

Comments
"Upgrading" is bound to be self-defeating.

Eventually complexity will get beyond anyone understanding the entirety. "Simple" parts come to be taken for utterly reliable yet almost anything more complex than a brick necessarily contains at least one undiscovered flaw. Huge automated systems can be brought down by failure of a single component, as I recall was the case with the classic New York blackout, and a more recent one. Then there's the Atlantic Airbus crash that seems to have been due to iced-over Pitot tubes. And if reports are correct that people slavishly follow GPS directions without regard to the road, then this constant "upgrading" is a clear and present danger.

posted by : bigger_luddite, 08 July 2010 Complain about this comment
aboutus
Advertisement
Subscribe to INQ newsletters
Advertisement
INQ Poll

Facebook starts selling shares

Will you buy Facebook shares?