Airline reservation systems built on more than half-century-old technology are increasingly showing cracks in the system. Delta’s computer meltdown, which was blamed on a small fire at the company’s Atlanta headquarters, grounded about 2,000 flights this week when backup power failed to fill the void left by the main computer system.
Delta has not released information detailing what went wrong, or how much it has invested in reservations technology over the past decade, but the company attempted to assure customers – and investors – that it is working to prevent future outages. "We have a new CIO who has a go-forward plan to ensure Delta is on the cutting edge of customer service technology while strengthening our IT infrastructure so that it is reliable, redundant and nimble," the company said in a statement.
Delta's Chief Executive Ed Bastian said in a video apology earlier this week that the airline has invested hundreds of millions of dollars on technology infrastructure upgrades and backup systems over last three years. "I'm sorry that it happened," he said of the power failure.
Industry experts say that the reservations systems of the biggest carriers – including American, Delta, United, and Southwest – mostly run on an IBM operating system called Transaction Processing Facility, or TPF, which was designed in the 1960s. IBM still updates the software, with the last major rewrite about a decade ago.
Meanwhile, airlines have piled special features allowing for mobile check-ins, seat selection, and cabin upgrades, on top of the TPF.
"They have surrounded that old industry infrastructure with modern technology," said Bob Edwards, who was United Continental Holdings Inc's chief information officer until 2014, in an interview with Reuters. "Those systems have to always reach back into the old core technologies to retrieve a reservation or to figure out who flies between Dallas and New York City."
Airlines, which are experiencing record profits after years of losses, are hesitant to make major network upgrades in fear that the technology could fail during the transition, says Gartner analyst Mark Jaggers. They’re afraid that they cannot afford to take them down to add equipment, install patches, and perform other maintenance, he says.
But not dealing with upgrades is also costing airlines – in some cases tens of millions of dollars in refunded tickets, missed bookings, canceled flights, vouchers, staff overtime, transportation, hotel and meal accommodations for stranded travelers and crew, and other expenses, as The Street points out.
When Southwest canceled 2,300 flights three weeks ago, also as a result of a system outage, that failure cost the company at least $54 million in lost revenue, according to figures that the company released. That could rise to $82 million at a final tally, the Street reported.
And it’s not just Delta and Southwest. JetBlue passengers faced major flight delays in January after a power failure, and then again in May. In the height of the summer travel season in July, United grounded its flights globally due to a router malfunction.
Some airlines are risking outages that might cost them $20 million to $40 million instead of investing $100 million on technology upgrades, Henry Harteveldt, founder of the travel consultancy Atmosphere Research Group, told Reuters. But despite record profits, it’s hard to convince airline executives to spend money on their back-end systems, says Mr. Edwards, formerly with United.
"When fuel prices are low and there's extra cash on hand, they want to spend it on the cool shiny things like planes and mobile apps," he told Reuters. "Nobody gets excited about the data center."
This report includes material from Reuters.