Airspeed sensors are what's crucial, not GPS which provides groundspeed. The airspeed sensors (pitot tubes) iced up.

To me, this is a bit more impressive...

The 747 that crosses rapidly before the American looks like Air France.

People have done all that here in the US, and aside from pissing off some controllers, being told to call phone numbers upon landing, and possibly getting an FAA-mandated vacation from flying, accidents aren't taking place. Have a look through some of the incident and ASRS reports, and I'm sure you'll find plenty of…

@TomXP411 , in most of the US you neither need to talk to ATC nor file a flight plan to fly, and our General Aviation numbers are orders of magnitude greater than those in China. Some safety comes from Big Sky Theory ( [en.wikipedia.org] ), but mostly it's because our system is designed from the ground up to prevent…

@matt.quadros , worst case scenario is a tire is blown, possibly puncturing a fuel tank, in the presence of brakes that exceed 1000c and the plane potentially burns. "Just" testing the brakes in a million pound airplane with tires at 200psi isn't the same as getting in your car and slamming on the brakes to see if you…

The tires on the Airbus are also supposed to, but the wheels weren't strong enough and they deformed when the tire pressure grew, failing before the heat plugs in the tires melted. [www.flightglobal.com]

The pilots are test pilots...it's their job to evaluate new systems and manage the risk associated with that. I'm sure they planned and simulated things as much as possible, but for FAA certification the test has to be performed in the actual aircraft. Recently, things went wrong on a test flight for Gulfstream's new…

@ thegamefreak0134: I don't think they've turned a 747-8 into an RC plane yet. Still pilots on board.

For the record, this (my video) is what's NOT supposed to happen... Airbus A340-600 rejected takeoff test. Summary: a fire starts, tires explode, crew has to evacuate plane.

@Smeghead: I don't disagree with anything you've written above. Redundancy and system logic is key. I can't dig through the 200+ page final report, but as I understand it the autothrottles were slaved to only one of the two radalts on board (the left side). The right autopilot (slaved to the right radalt) was engaged…

@vc-10 it's precisely Alpha Floor Protection that played a critical role in the Air France 296 disaster. The computers do not always give the expected response when outside the standard flight envelope, and the reliance on automation causes pilots to expect one thing from a system but they're given another. The Airbus…

I don't pilot a helicopter (but hold a fixed-wing pilot's certificate). I work as an aerial technician, working with aerial photographers and video crews. I ride in the back of the helicopter and assist, and liase between photog and pilot as well as bring local knowledge for planning aerial shoots.

Image #1 in the gallery is Manhattan. In fact. there's only one night shot over London, and that's in the main article, not in the gallery.

What about scenarios when computers are fed data from faulty sensors, like the recent Turkish 1951 crash with a bad radio altimeter that jumped from reading 1950' to -8' on the approach, retarding the autothrottles? Interesting read.. [aviationknowledge.wikidot.com]

For your morbid interest:

Gizmodo's comment system won't let me post the chain of links, but basically search for "uncontained engine failure passenger -a380" took me to an airliners dot net page, which led me to search for "National Airlines DC-10", and I found it on that wikipedia page.

Did some digging and found it, you're correct: [en.wikipedia.org] (National Airlines Flight 27, a DC-10). There are plenty of things a pilot can do that won't end well for all aboard.

Here's a more balanced article from Reuters: [www.reuters.com]

The picture at the lead of this article is a lexan tank filled with seawater used to transport the CVR and DFDR, both of which are bright orange.