Discuss This Article

Hearing Voices

Part science, part art and part human relations, Anna Cushman's job as a cockpit voice recorder analyst is to help find out what went wrong, and, maybe, keep it from happening again.

By Ray Bert '93

"The first thing that struck me was the devastation," says Anna Cushman '91 of her first look at the Pentagon on the afternoon of Sept. 11. "The TV shots I'd seen really didn't give you a sense of the magnitude of the destruction.

"The Pentagon is a massive, solid building, and the gaping hole where the roof had collapsed was mind-boggling. The building was completely disintegrated inside. There were piles of debris several feet deep on the ground floor, and where there wasn't debris, there was about half a foot of water or sludge."

Somewhere in that massive pile of rubble lay two mangled metal containers that might reveal what happened aboard American Airlines Flight 77 in the minutes before terrorists crashed it into America's military headquarters. As a cockpit voice recorder analyst for the National Transportation Safety Board (NTSB), it was Cushman's job to help locate the airplane's black boxes, as the voice and data recorders that all airliners carry are known informally. It was the first crash site she'd visited.

Over the next few days, working the 3 p.m. to morning shift, she and several other NTSB experts struggled to separate airplane parts from office parts. Early on the morning of Sept.14, while Cushman was at the site, the cockpit voice recorder, or CVR, was found. It was quickly transported across the Potomac to the NTSB lab in Washington, D.C., where Cushman works with three other analysts, and its data was downloaded.

Ordinarily, that would have been just the start of Cushman's association with the device, but this time, it was the end. The events of Sept. 11 had already been classified as criminal acts, rather than accidents, so the FBI, which has its own forensic audio lab, took charge of the box and its data.

That's also why Cushman can't say much more about her role in that investigation, or about the work she did on the recorders recovered from Flight 93, which plowed into a field in Pennsylvania after passengers apparently thwarted another hijacking. Like the Pentagon CVR, the black box from that plane came to NTSB only for the extraction of its data before being turned over to the FBI. The recorders from the two planes that struck the World Trade Center have yet to be found.

Incidents and Accidents

The air disasters of Sept. 11 were anomalies. Ordinarily, the cause of a crash or other aviation incident remains at least something of a mystery until the NTSB conducts its investigation. And for most of the 4,000 aviation incidents and accidents the agency investigates in a typical year, the mystery can be solved without consulting the CVR. "The investigator in charge determines whether it's necessary to download the information," Cushman says. "If you have a good pilot interview and it's obvious what happened, you might not need to."

Cushman's group sees an average of one recorder per week, "though they always seem to come in five at a time," she says. Most come from smaller commercial, corporate and private jets, typically involved in relatively minor events like runway overruns. "Most of the incidents we get don't make the front page of The Washington Post--they don't make the Post at all," Cushman says.

Once a CVR has been delivered to NTSB, Cushman begins work immediately, day or night. Depending on how badly the unit has been damaged, she may have to cut the box open to get at the tape or the memory chip. She downloads the audio information and prepares a sound spectrum analysis and a transcript. The transcript, in whole or in part, may be released at public hearings and in NTSB reports, but because of the sensitive nature of the sounds they contain, Congress prohibits the NTSB from releasing the actual tapes, themselves.

The transcript is prepared by a group, led by Cushman, that includes representatives from the Federal Aviation Administration, the airline involved, the airplane and engine manufacturers, and the pilots union. The process can be tedious: Cushman says a 30-minute recording can take a day or more to transcribe, due to constant rechecking and the subjective nature of trying to discern words spoken in the loud cockpit environment. For a serious accident, "just the last 30 seconds of the recording can take an hour to do," she says.

Running a CVR meeting is an exercise in group dynamics, which is why all CVR analysts must also be pilots. "It's hard to get a group of pilots to work together when they think you don't know anything about flying," she says. "For instance, if you've never experienced it, you'll have a hard time understanding how a pilot can be upside down in the clouds and not 'feel' upside down. Someone without pilot experience is at risk of being run by the group, instead of running the group."

Cushman's technical expertise comes into play during the sound spectrum analysis, which creates a set of computer-generated waveforms (amplitude vs. time) and spectrograms (frequency vs. time vs. amplitude) that turn the audio information into three-dimensional pictures and help her identify the likely source of individual noises. There's still as much art as science to the process, she says. "You may identify the sound of a hydraulic pump amid the noise on a Learjet. But on the next Learjet you do, even if it's a sister ship, that sound might not record the same way because the mike might be older."

Learning to Fly

"I've always been interested in airplanes, and I'd always wanted to learn how to fly," Cushman says. Despite her interest in aviation, Cushman passed up the Air Force Academy, where she was also accepted, to attend WPI. "I chose WPI because of the projects. And all of my projects were really cool. I ended up doing two projects for NASA, and my Sufficiency was on photography in flight. Those projects got me co-op jobs at Textron Lycoming."

She played on the tennis team ("I probably hold the losingest record at WPI," she says. "I can count on one hand the number of times I won in four years.") and was a member of Alpha Gamma Delta sorority. With a degree in mechanical engineering, she graduated in 1991, a low point for the aerospace industry. Facing a dismal job market, she opted to accept a scholarship and earn a master's degree at Tufts University. In 1993, she found a job at Sikorsky Aircraft in Stratford, Conn., teaching helicopter ground school hopefuls about hydraulic and electrical systems. She moved on to do engineering work, including waveform analysis in radar cross-section studies, similar to her current sound spectrum work at NTSB.

Sikorsky had one especially enticing perk: an education benefit that enabled her to realize her longstanding ambition to earn her pilot's license. The program produced an unexpected bonus: Cushman met her husband, Jan Fredrik Wold, while at flight school. Actually, she admits somewhat sheepishly, he was her instructor. "We didn't start dating until he was no longer my instructor--that should be made clear!" she adds with a laugh.

As a pilot in training, and one whose husband flies planes for a living, Cushman found herself developing an interest in flight safety. Checking out the NTSB Web site one day, she saw a posting for a CVR analyst, and after some internal debate, she applied just before the closing date. "When the offer came my way, I couldn't refuse it," she says.

Ironically, her husband's occupation, which was one of the reasons she ended up at NTSB, now determines which incidents she can investigate. Because he flies for American Eagle, the NTSB requires that Cushman recuse herself from any incident involving American Airlines (including last November's crash of American Airlines flight 587 in New York) because of the potential for a conflict of interest.

The Human Aspect

Concerns of a different type spring to mind when many people contemplate what Cushman does for a living. "Isn't it depressing?" they ask.

"Most of the stuff we do isn't as morbid as it sounds," she says. "But, yes, on occasion, it can be what you'd expect, what the general public thinks that we do all the time, which is listening to people die.

"There really isn't any way to train for that part of the job," she continues. "I've done several fatal accidents, and I can't say that you get immune to them, because that's not how it works. But most of the time, because of the actions of the crew, you're able to do your job because they were doing theirs."

The human aspect of voices recorded on tape, along with the potential for those tapes to hold clues that may help solve an aviation mystery, makes the cockpit voice recorder a sensational part of any accident investigation, even within the NTSB. Cushman says of co-workers, "If they're not looking over your shoulder, they're poking their heads in the lab every two seconds, wondering where you're at."

Though it's not uncommon for the media to camp out at the NTSB when they know a CVR has arrived, Cushman says it's important to understand that the in-flight recording is not the last word in most accident investigations. That's largely due, she says, to the subjective nature of her job. "The CVR might point the investigation in a particular direction," she says, "but it might turn out to be the wrong direction. Because of that, what's on the recorder is considered secondary supporting evidence in an investigation."

Still, she says she never loses sight of the importance of the work she does, and its capacity to provide answers and, possibly, prevent future accidents. Nor, she says, can she rid herself of the memories of those haunting voices and telltale sounds contained on the tapes of those rare and tragic accidents. "That doesn't go away," she says. "Not ever, I think."

-Bert is a freelance writer living in Virginia.

Related Links

Maintained by: webmaster@wpi.edu
Last modified: Sep 15, 2004, 12:57 EDT
[WPI] [Alumni] [Home]