News

‘Drone With Ears’ Uses MEMS Microphones to Hear Disaster Victims

December 13, 2023 by Aaron Carman

You’ve heard of eye in the sky, but this new drone uses hearing to locate disaster victims.

Researchers at the Fraunhofer Institute for Communication, Information Processing, and Ergonomics (FKIE) have developed a “drone with ears” that can listen to the cries of disaster victims trapped under rubble. This capability may be especially useful in complex urban environments with multiple reflections and obstacles.

 

A new FKIE drone

A new FKIE drone uses a MEMS microphone array to locate sounds, allowing disaster victims to be found even if they are covered in rubble. 
 

After a disaster, rescue teams often take to the skies to locate victims using manned helicopters or unmanned drones, each of which relies heavily on vision technology. Emergency drones typically use optical sensors such as daylight or thermal cameras to sense humans. If, however, a victim is trapped under rubble, vision alone is not enough to locate them. In addition, adverse weather conditions can prevent effective operation.

FKIE researchers set out to make drones with a new microphone array, so the device could both hear and see their targets.

 

Keeping an Ear Out

To give rescue teams a way of listening below the surface, FKIE researchers have leveraged a MEMS microphone array to not only hear victims but also determine the location of their calls for help. The array can be carried by a drone and use AI signal processing to separate screams or claps from ambient noise.

 

Test drones from FKIE have used only MEMS arrays

Test drones from FKIE have used only MEMS arrays to locate the sources of sounds, improving the performance of disaster recovery drones. 
 

The FKIE drone, dubbed LUCY (Listening System Using a Crow’s Nest arraY), leverages a unique MEMS microphone array to localize the source of sounds. The crow’s nest array, also used in microwave systems, enables beamforming and incidence angle estimation using 48 independent microphones. In addition, the semi-random layout of the array minimizes grating lobes, which could impact the final operation of the drone.

 

The LUCY MEMS array

The LUCY MEMS array enables beamforming to localize the source of sounds, providing a new sense for disaster recovery drones. 
 

To separate human cries from ambient sounds, LUCY leverages AI-enabled signal processing. As such, pilots can filter out wind, propellor blades, and natural sounds and focus solely on noises coming from trapped humans. This feature opens the possibility of automated swarms of LUCY drones in the future.

 

LUCY Complements Other Imaging Tech

FKIE reports that LUCY can isolate impulse sounds with high precision. Despite its promising outlook, LUCY is not meant to be a standalone system. Rather, it is meant to be integrated with state-of-the-art cameras and thermal imaging technology to provide more resources to first responders. In addition, since LUCY can be integrated directly into commercial drones, designers can leverage the technology in their own systems, perhaps in new applications outside disaster recovery.

With LUCY, trapped victims can now alert rescue teams without needing to see them, increasing the probability of rescue in the crucial minutes following a disaster.

 


 

All images used courtesy of FKIE.