Drones May Soon Avoid Obstacles Like Bats Using Echolocation
Dhir Acharya - Feb 15, 2020
Scientists have come up with a way to help drones see the world like humans do, using a technique called echolocation found in bats.
- 22 Most Innovative Gadgets Of The Last Decade
- Baghdad Drone Strike Killed Iran General Qassem Soleimani
- This Drone Will Use Special Seeds To Plant 1 Billion Trees In 8 Years
With technological advancements, machines have come a really long way since they were first created centuries ago. However, they are still now so far from seeing the world like a real human, even navigation remains a challenge for machines. However, scientists have demonstrated a new way to improve navigation for drones using bat-like echolocation.
According to the study, the scientists, using a mathematical model, investigated how drones could rely on sound waves’ reflection to pinpoint the locations of different walls inside a room. To do this, they imagined a scenario where they put a loudspeaker in a room and equipped a drone with four microphones, putting it a random position in the room.
As stated by the paper’s co-author Mireille Boutin, with this configuration, the drone could measure how long it took a sound to be bounced back to the drone from a wall, then it could determine the distance from it to the wall as well as the walls’ configuration. Boutin said:
The drone imagined that the sound data comes from a virtual point behind the real walls rather than a speaker. Then, it uses the virtual points to compare the distance between the microphones and the virtual source then reconstruct 4 points on the wall.
After collecting that data, the scientists told the walls from one another using an echo sorting algorithm.
According to the authors, this approach, including equipping the drones with microphones and using echo sorting, helps solve a problem with previous drone models: detecting ghost walls. The authors placed microphones around the room, not at one point on the drone and found that the sound wave’s reflection could bounce back from an imaginary source inside the room. As a result, the drone could incorrectly believe that another ghost wall existed inside the room and it would visualize the room incorrectly.
The study is still in theory only but this technology can be applied successfully in reality while we will no longer need additional info to visualize the walls. The study’s authors also write that the applications of this technology can go beyond drones. Professor Gregor Kemper of the Mathematics Department at Technische Universität München said: