X

Tech beyond black boxes? It just won't fly

In a digital era where airplanes are jammed with phones and Wi-Fi, why are flights still relying on older technology to recover data? Photos: Retrieving data from the cockpit

Declan McCullagh Former Senior Writer
Declan McCullagh is the chief political correspondent for CNET. You can e-mail him or follow him on Twitter as declanm. Declan previously was a reporter for Time and the Washington bureau chief for Wired and wrote the Taking Liberties section and Other People's Money column for CBS News' Web site.
Declan McCullagh
7 min read
Greek investigators struggling to determine the cause of this week's deadly airplane crash have encountered an unexpected setback: One of the black boxes with vital flight data has not been recovered, and the other may be damaged.

A second problem in the search for clues in the mountains north of Athens lies in the design of the separate cockpit voice recorder, which stores only the last 30 minutes of audio. Because the Helios Airways flight apparently flew for hours after the pilots were disabled, the most valuable information may have been overwritten.

In a digital era where airplanes are jammed with telephones, satellite TV and Wi-Fi wireless links, relying on physical black boxes that can be lost or incinerated seems almost outmoded. So why don't commercial flights beam important data back to the ground every few microseconds?

Krishna Kavi, chairman of the computer science and engineering department at the University of North Texas, says that piggybacking an air-to-ground connection on top of an existing transmission link is feasible.

"You don't need a very expensive communication network to relay information from plane to ground," said Kavi, who has written about black box design. "If you can use a cell phone, recorders should be able to transmit that information to an airline company."

There are good reasons for air-to-ground streaming. If an in-flight recorder is lost or destroyed, the additional precaution of backing up flight data on the ground would help investigators reconstruct what happened.

An air-to-ground link also would permit information to be retained much longer; digital flight data recorders on current Boeing planes store data for 25 hours, and cockpit voice recorders store between 30 minutes and 2 hours. And searching for black boxes with damaged radio beacons is frustrating and expensive.

The federal government has never required air-to-ground data feeds, primarily because officials believe the cost of a reliable system remains too high.

"We feel our money's better spent making the recorders so they will survive, they can be found, and provide the economically reliable data that we need," said James Cash, chief of the vehicle recorder division at the National Transportation Safety Board. "I don't think you could ever make the economic justification that a real-time download is economically feasible to impose on the industry."

Other reasons include the challenge of engineering a system that could transmit even when the plane is in an unusual flight attitude--such as inverted or in a dive--instead of in normal flight. The NTSB says it's aware of only three crashes, not counting the Sept. 11, 2001 attacks, in which the flight data recorders were not recovered.

Really orange, not black
Dave Warren's interest in recording music led to the creation of the original black box in the mid-1950s.

An engineer at the Aeronautical Research Laboratory in Melbourne, Australia, Warren was helping to investigate the cause of an airplane crash that had no witnesses. He realized that reconstructing what happened would be easier if the information was stored onboard.

"Dave then recalled the world's first miniature (tape) recorder that he had recently seen at a trade fair," according to a history prepared by the Australian government. "Suddenly he could visualize such a recorder placed in all aircraft, continually recording details and able to be recovered after a crash."

black boxes

Today's black boxes--a term that encompasses the flight data recorder and the cockpit voice recorder--are required on commercial airliners and regulated by the Federal Aviation Administration and the European Organization for Civil Aviation Equipment. To aid in recovery, both are painted orange.

They're designed to survive the fiery temperatures and intense pressures of a crash. Regulations require that the boxes be resistant to 5,000 pounds of pressure, temperatures of 2,000 degrees Fahrenheit, and deep sea pressure of 20,000 feet. (Similar black boxes are popping up in cars too.)

Flight data recorders are plugged into the aircraft's main data bus, which lets them record up to 88 different forms of information such as altitude, airspeed, and the position of the stick used by pilots to control the plane's elevator and ailerons.

The data they safeguard can yield explanations that can help prevent future crashes. A January 2000 NTSB report on an Alaska Airlines flight that crashed off the coast of California determined, largely based on black box data, that the horizontal stabilizer on the tail was

jammed up, causing the nose to pitch uncontrollably downward.

"Instant messaging" for pilots
Many airliners already transmit a small fraction of this information to a network of hundreds of ground stations around the world through a digital data link called ACARS, or Aircraft Communication Addressing and Reporting System. (ADS-B is another data link with similar characteristics.)

The two-way ACARS link, which works over VHF radio, resembles a kind of slow text messaging for airplanes. It lets airlines remotely request information such as fuel consumption, engine status and landing gear position. Pilots can ask for text-based weather reports from the ground.

But ACARS is designed for short bursts of information--not a constant stream of data--and is typically used only a few times during a normal flight. Revamping the low-speed ACARS system to handle the volume of information retained by a black box would be a daunting task.

"You don't need a very expensive communication network to relay information from plane to ground."
--Krishna Kavi, chairman of the computer science and engineering department, University of North Texas

Satellite links could handle the necessary bandwidth--88 data points, plus multiple audio streams stored by the cockpit voice recorder. But the cost for installation and usage makes them by far the most expensive alternative.

That's why black boxes are still the best option, says Cash, the NTSB division chief. "Who's going to store that data? Where's it all going to go? Most of it would be totally routine--99.9 percent of it would be totally routine all the time," he said. "Recording onboard is much more cost-effective and simple and reliable, knowing that you have to sometimes pay the price to find it."

One reason to maintain the current system, the NTSB believes, is the newer breed of black boxes that use solid-state memory are more likely to remain intact after a crash. Older black boxes relied on more delicate magnetic tape. (Seventy percent of U.S. commercial flights have switched to solid-state recorders.)

A bill introduced in June would require a duplicate set of black boxes that are designed to separate from the airplane at impact. That would, its proponents believe, help avoid a repeat of what happened on Sept. 11, 2001, in which flight data recorders were recovered from only two of the four hijacked planes that crashed.

Kavi, the computer science professor, has proposed a hybrid approach in which only unusual airplane data would be transmitted to the ground. "You don't really have to send all the data," he said. "What you can do is send only the data when it falls outside of normal range, so if your altitude is as it is expected, you don't need to send the data, if you think the altitude is not in the proper range, (then) transmit."

A paper Kavi co-authored titled "Glass-Box: An intelligent flight data recorder and real-time monitoring system" goes even further, however. The paper suggests permanently recording all the data that's currently stored and discarded. Once that's done, the data "collected from numerous flights can be correlated and data-mined to construct scenarios that could lead to unsafe incidents."

In such a scenario, a software program running either on the plane or the ground would analyze the data flow and detect potentially unsafe situations that a human pilot could not.

Privacy questions: a pilot's last words
Complicating the question of beaming cockpit audio to the ground is the privacy of audio recordings. Under federal law, the NTSB "may not disclose publicly" any audio from a voice recording--only an excerpt from a transcript may be released.

Computer hobbyists with radio receivers regularly capture ACARS transmissions between planes and the ground, and shareware programs with names like DACARS and KRACARS can decode the data stream. Unless the data were securely encrypted, the prospect of voice transmissions would likely encounter stiff opposition from pilots and airlines.

Airline pilots have opposed, for instance, an NTSB proposal to place cameras in cockpits and save captured video streams to the black boxes.

Still, as the price of data transmission gradually falls, industry groups expect such air-to-ground links to become more popular. Interest in technological upgrades could be spurred if Greek investigators are unable to retrieve valuable data that could have explained what killed 121 people.

"Generally speaking, real-time data link technology is a good idea," said Matt Grimison, a spokesman for the Aerospace Industries Association, which represents companies including Boeing, Northrop Grumman and Aerojet. "It holds a lot of potential...The development of the technology isn't quite where it needs to be right now in order to make it a reality, but clearly it is the technology of the future."

CNET News.com's Anne Broache contributed to this report from Washington.