SCOTT SIMON, HOST:
Every year, the Coast Guard responds to more than 16,000 distress calls, but about 150 of those calls are frauds. Those hoaxes can drain the Coast Guard of valuable resources and put crews out of position when someone who's really in distress needs help. Now, the Coast Guard looks to voice forensic technology for help. Al Arsenault heads the Coast Guard team that tries to detect fraudulent distress calls. He joins us now from New London, Conn. How are these fraudulent calls so dangerous?
AL ARSENAULT: It takes away valuable Coast Guard resources that could be used for real search and rescue cases. And one of the major things is it puts our people at a significant risk.
SIMON: How would voice profiling help?
ARSENAULT: Here at the Coast Guard Research & Development Center, we actually looked at several technologies. We have - we've teamed up with Carnegie Mellon University. They have really helped us basically determine if it's the same caller that's making these calls and then gives us some information to help try to track down these people.
SIMON: What kind of information can it give you?
ARSENAULT: As far as the forensics goes, the Carnegie Mellon team can actually tell us the age of the person, weight, height, country of origin. And they can also give us some information on the surroundings the calls were made from. For example, is there a lot of concrete in the area? Are there a lot of windows? So they can give us a lot of information.
SIMON: Mercy, an area like that could be in Oregon or it could be in Maine. How do you know the difference?
ARSENAULT: Exactly. The other tools that we're employing in addition to the voice forensics is some direction-finding capability, which can actually give us the direction the calls are coming from. We're also using some social media analytics tools because what we have found out based on years of experience is that these callers like to brag on social media. So they might like to see a Coast Guard response - whether it's a ship, a boat, an aircraft - and post pictures online.
SIMON: This going is to sound like an awfully naive question, but obviously some voices are pretty easy to imitate. Can voice forensics be fooled, too?
ARSENAULT: It's a lot harder. They're - and I want to say now, I'm certainly not the expert.
ARSENAULT: Basically, everybody's windpipe acts differently than somebody else's. So even though they try to imitate, their windpipe responds differently when they're actually talking. For example, Sometimes these people are at different levels of intoxication. And we're able to pinpoint the calls to the same person even though, you know, if I was listening to him, I would say there's no way that's the same person.
SIMON: You know, I've got to tell you, I'm impressed by what you say the forensic technology can do and also a little intimidated.
ARSENAULT: Yes. And, you know, this - the forensics piece, it's one part of building a case file, kind of weird using DNA as evidence was several years ago in court. Right now, we're using that as one part of the evidence package, but it's becoming more and more reliable as time goes on.
SIMON: Al Arsenault of the Coast Guard's Research & Development Center. Good luck to you, sir.
ARSENAULT: Thank you very much.
NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.