United Airlines computer glitch creates cascade of delays

A computer glitch grounded all United domestic flights for about two hours on Sunday evening, causing further delays to ripple across the United States.

First Look

United Airlines grounded all domestic flights for about two hours on Sunday evening, after a computer glitch struck their system.

Leading to six cancellations and 200 delays, it was the latest IT issue to cause the airline problems in recent months. And United is not the only carrier to have suffered such setbacks of late, as the complexity of the computer systems in charge make it all but impossible to proceed when they malfunction.

On this occasion, it was only internal US flights that were affected, and the company has implemented a travel waiver whereby passengers can rebook flights for no extra charge, provided the class, city of departure, and destination all remain the same.

Just last October, a similar problem caused delays to United flights worldwide. That followed an incident in July, when some 4,900 flights globally were grounded. Less than two months earlier, all takeoffs in the United States were halted after computer issues.

And United is not unique in wrestling with recent IT glitches. Air Canada and Porter Airlines (based in Toronto) were forced to cancel flights due to computer malfunctions just last week. Delta Air Lines had to cancel some 2,000 flights in August, after a power outage incapacitated computer systems. British Airways experienced similar disruption the following month.

So, what’s to be done?

It seems unlikely that airlines are going to abandon computer systems anytime soon, but researchers continue to develop means of more accurately predicting delays, and of being able to inform customers more quickly about their predicament and decreasing confusion – and frustrations – about the cause of delays.

Researchers at Binghamton University in New York, for example, announced in November that they had devised a new computer model that can “more accurately predict delays faster than anything currently in use.”

The model can better account for qualitative factors, like the weather, instead of numerical variables, using past flights' data to predict future hiccups. 

The main thrust of the project, study lead author Sina Khanmohammadi said in a press release, was “to provide more accurate delay information to the customers, and hence gain customer loyalty.”

This report contains material from Reuters and the Associated Press.