WSJ: Incidents Prompt New Scrutiny Of Airplane Software Glitches

[question everything]

The Wall Street Journal

Incidents Prompt New Scrutiny Of Airplane Software Glitches

As Programs Grow Complex, Bugs Are Hard to Detect; A Jet's Roller-Coaster Ride

Teaching Pilots to Get Control
By DANIEL MICHAELS and ANDY PASZTOR
May 30, 2006; Page A1

As a Malaysia Airlines jetliner cruised from Perth, Australia, to Kuala Lumpur, Malaysia, one evening last August, it suddenly took on a mind of its own and zoomed 3,000 feet upward. The captain disconnected the autopilot and pointed the Boeing 777's nose down to avoid stalling, but was jerked into a steep dive. He throttled back sharply on both engines, trying to slow the plane. Instead, the jet raced into another climb. The crew eventually regained control and manually flew their 177 passengers safely back to Australia.

Investigators quickly discovered the reason for the plane's roller-coaster ride 38,000 feet above the Indian Ocean. A defective software program had provided incorrect data about the aircraft's speed and acceleration, confusing flight computers. The computers had also failed, at first, to respond to the pilot's commands. Within weeks Boeing Co. warned airlines world-wide to install a fix provided by Honeywell International Inc., which makes the flight computers and supplied the faulty software.

(snip)

In theory, most advanced jetliners can take off, climb, navigate along a prescribed route, descend to their destination and roll to a halt at the end of the runway -- all without human intervention. Autopilot programs were first created to make planes fly more smoothly and reduce pilot distractions by taking over rote tasks. Today's software also handles many other vital aspects of flight, such as adjusting cabin air pressure, maximizing fuel efficiency and warning of impending mechanical breakdowns or collision threats. Advances planned for the Airbus A380 superjumbo jet, due to start passenger service later this year, and Boeing's long-range 787 Dreamliner, due in 2008, will take automation to new heights. Instead of independent hardware and software systems for each task, the new jets will save weight by relying on redundant central computers to run the whole plane. The systems will have safeguards to prevent programs from interfering with and confusing each other. For example, if speed readings from sensors differ widely, the computers are designed to disregard the most extreme measurement and try to figure out which reading is correct.

(snip)

Nonetheless, the design of computer systems intended to eliminate cockpit mistakes carry their own hidden risks of human error. In February 2005, multiple computers meant to back each other up mistakenly cut the flow of fuel to two of the four engines on a packed Virgin Atlantic Airbus A340 headed from Hong Kong to London. The crew made an emergency landing at Schiphol Airport outside Amsterdam. British investigators have recommended design changes, including better warnings on fuel levels. Virgin Atlantic says it is cooperating with the investigation and the safety of the passengers was never threatened. Last October, a 90-second computer hiccup aboard a British Airways Airbus A319 on a night flight from London to Budapest temporarily shut off nearly all the cockpit lights and electronic displays, along with radios and autopilot systems. British investigators say they have unearthed four similar cases on Airbus jets.

(snip)

Malaysia Airlines Flight 124 is a case in point. Boeing's 777 jets started service in 1995 and had never experienced a similar emergency before. According to Boeing and Honeywell, the source of the problem was a revised computer program that had recently been installed on all 777s to fix a minor navigation flaw. Honeywell and Boeing didn't know that the new program had a defect: It simultaneously told the autopilot that the plane was flying dangerously slow and much too fast. Investigators are still trying to figure out what circumstances triggered the program to give the conflicting data. Flight computers struggled to respond, and the plane's speed plunged to 150 knots from 270 knots. The control column shook to warn of an impending stall. Soon after the incident, Boeing issued a safety alert advising that, in such circumstances, pilots should immediately disconnect the autopilot and might need to exert an unusually strong force on the controls for as long as two minutes to regain normal flight.

(snip)

URL for this article:
http://online.wsj.com/article/SB114895279859065931.html (subscription)