Bat drone orients itself by sound

February 15, 2023

Smart EPFL development uses low-cost microphones and a piezoelectric buzzer

The new “Crazyflie” drone developed by researchers at the Swiss Federal Institute of Technology Lausanne (https://www.epfl.ch) (EPFL) has neither a camera nor LiDAR or radar. Nevertheless, it flies even at night without touching obstacles. The model is the bat, which has the same capabilities. It emits ultrasonic pulses and uses the reflected signals to detect obstacles in a flash, which it avoids, or prey objects, which it greedily pounces on.

Four microphones and a buzzer

The drone is equipped with a piezoelectric buzzer mounted in the centre of the flying object. So-called MEMS microphones are installed at the end of the four arms, which are located between the four webs with the drive motors, to pick up the reflections of the buzzing sounds that are emitted with short interruptions. From this, the microcontroller on board creates an image of the surroundings so that the drone flies through the three-dimensional course without visual control.

MEMS microphones are miniaturised microphones designed in SMD technology for direct use on electronic circuit boards. They are particularly inexpensive, as are the other components of the drone. “We deliberately did without expensive measurement microphones and loudspeakers in order to keep the costs of the drone low. If there is a strong reflector like a wall nearby, the sound is reflected and interacts with the direct sound at each microphone,” says EPFL developer Frederike Dümbgen

Accurate acoustic localisation

Depending on the distance and frequency, these two signals weaken or amplify each other. From this, the microcontroller determines the distance of the obstacle. Since there are several microphones, it is also possible to determine the angle at which the reflector is in relation to the drone. Dümbgen outlines the future of the Crazyflie as follows: “We humans rely more on our ears when it is dark and on our eyes when there is loud ambient noise. I believe that truly intelligent robots should have this ability as well.”

Related Articles

Critical Communications: From Radio Call to Operational Record

Body-worn cameras are sending live video to control rooms, satellites are extending connectivity beyond terrestrial coverage, and AI systems are beginning to log decisions made in emergency centres. Critical communications is expanding beyond voice radio. The...

Biometrics: From Face Matching to Trust Infrastructure

Facial recognition is being used to speed up border crossings, reinforce passkeys and support digital financial services. At the same time, companies are linking biometric data with access rights, risk assessments and government screening systems. Developments in the...

Share This