If there is a building destroyed, or a catastrophe such as an earthquake people could be trapped under the rubble, rescue crews need to react really quickly.
The use of drones in urgent situations is becoming more popular as the technology gets increasingly refined. Now, a team of researchers has developed a system that could fly over disaster areas and help pinpoint the sounds of people who are screaming for help.
While the need for such an invention is grim, recent years have demonstrated how useful such technology could be. Particularly in communities that are prone to wildfires, hurricanes, floods, and other forms of extreme weather that could seriously benefit from scream-identifying drones. These drones could be essential in the coming years as climate change is only worsening storms and fires. They would also be valuable for wilderness search and rescue missions.
Macarena Varela, a researcher at Fraunhofer FKIE, who worked on the array and presented it to the Acoustical Society of America’s annual conference last Tuesday said, “If there is a building that is being destroyed, or if there is a catastrophe such as an earthquake people could be trapped under the rubble, rescue crews need to react really quickly.” She continues, saying, “It’s difficult for them to find the location of these people, but we could easily assist them in finding these locations if we’re using a UAV flying over a big area detecting human screams and other noises.”
The team created the scream-identifying drone by assembling a database of screams (think Monster’s Inc.) and other “impulsive” sounds people make when they’re trying to get the attention of rescuers, such as tapping, clapping, and stomping. They then applied deep learning techniques to test the database with a microphone array set up inside a lab to see if the system could accurately distinguish between human-in-trouble sounds and other types of noise.
According to Varela, the team found that the lag between the microphone system picking up the sound of a scream and accurately identifying its location was only a couple of seconds. The team was also able to run speech clarification tests, which allowed them to amplify the sound of a distant mumble and refine it until it was a clear sentence. This technology could help rescuers get important information about the specific needs of people who are trapped or in trouble.
Varela and the rest of the team then placed more than 60 digital microphones on top of a drone to test the system outside of the lab. Digital microphones don’t need a sound card and aren’t as bulky as traditional microphones, making them easier to deploy on drones. The microphones themselves aren’t very expensive either, costing them just a few Euros each, and it would be easy to add them on to drones that are already being used by first responders to get aerial data of disaster sites.
The idea would be for the drones to fly outdoors and pinpoint noises coming from collapsed buildings, then fly near where the noise source is and hover above it to identify the position, but the technology could be applied in many other situations. It can even be used by a rescuer carrying it on the ground or wherever they are needed, as it can help them identify sounds that human ears miss.
Varela says that they haven’t been able to test the mobile drone system in a real-world situation, but suspects that a tool like a drone could be extremely useful to rescuers who need to cover large areas as quickly as possible. Varela hopes that more sensors like these ones could be used to locate unconscious victims as well to make the operation even more comprehensive.