Home https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ Science https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ Scientists teach drones to hear your screams

Scientists teach drones to hear your screams



An illustration for an article titled Drone Screening Detectors can save you

Photo: Gary Broome (AP)

Screaming in the void may be a thing of the past. A team of researchers said this week that it had developed a system that would allow rescuers to fly drones over distressed areas and identify the sounds of trapped people screaming for help.

The new equipment was presented on Tuesday at the annual conference of the Acoustic Society of America. “If there is a building that is collapsing or there is a catastrophe, such as an earthquake, people can fall into the trap under the rubble, rescue teams must react very quickly,” said Macarena Varela, a researcher at Fraunhofer FKIE who worked on the massif and introduced him of the Acoustic Society over Zoom. “It’s hard for them to find the whereabouts of these people, but we could easily help them find them if we used a UAV flying over a large area, detecting human screams and other noises.”

As grim as it sounds, the last few years have shown many cases of using this type of technology. Communities are built in places prone to fires, hurricanes, floods and other forms of extreme weather. There were often people in areas affected by hurricanes or forest fires who needed help detached from cellular service or other forms of communication. Horrifying images often appear after natural disasters such as Camp Fire or Hurricane Harvey of rescuers working around the clock to find survivors.

Climate change is exacerbating storms and fires, making the need for life-saving technologies even more urgent. Inventing ways to use technology to reduce the time it takes to find people seems like a solid choice, even when it comes to something as sinister as a drone that knows your screams.

The process of creating this screaming drone began by collecting a database of screams (wow) and other “impulsive” sounds that people make when they try to get the rescuers’ attention, such as tapping, cracking and knocking. Using deep learning techniques, this database was tested with a microphone grid set up in a laboratory to see if the system could accurately distinguish between these sounds in people in need and other types of noise. The delay between the microphone system, which captures the sound of screams and accurately identifies its location to rescuers, Varela said, is only a few seconds. Researchers also conducted tests to clarify speech, which allowed them to amplify the sound of distant muttering – like someone who hears from afar – to a crystal clear sentence. This can help rescuers identify important information about the specific needs of people who are trapped or otherwise in distress.

Varela and her colleagues then placed more than 60 digital microphones on a drone to take the system out of the lab and into the (literal) field. Unlike traditional microphones, digital ones do not require a sound card and are much less bulky, making them easier to deploy drones. Researchers say the microphones themselves aren’t that expensive for just a few euros per pop, while the first responders is already beginning to use drones to obtain aerial data on disaster sites, so it can be an easy addition.

“Our idea would be to fly outdoors and detect noise coming from collapsed buildings, but this array can be used in many other situations and used by the person carrying it,” she told a news conference. “Drones can cover much larger areas faster than humans on earth, but of course they could also be used by rescuers who may not hear the noise and who could walk with it to be able to the system to detect this noise. “

IN video of the drone setup posted on the website of the Acoustic Society of America, you can see the drone in action aimed in the middle of a flight at a researcher screaming for help. “The idea would be to use a drone and fly in the area as a whole to pick up the sound, and as soon as the sound is detected, the drone will fly close to the place and bend down to identify the exact position it is coming from. the sound, ”Varela said.

While larger arrays work great in the lab, Varela said they have not yet fully tested the mobile system for real-life drones. Still, she expects that a tool like a drone can be very useful for rescuers working to cover large areas in a short period of time. Other sensors, such as those that can be used to detect unconscious victims, may also be included to aid in a more complete operation.

“When a disaster strikes, every minute counts for finding missing people,” Varela said.


Source link