Drone guidance system uses sound to "see" walls
Many drones already utilize ultrasound to detect individual obstacles lying directly in their path. A proposed new drone system, however, would use sound to determine where all of the walls are within a room – providing guidance when vision- or light-based systems could not.
The research is being led by Prof. Mireille "Mimi" Boutin of Indiana's Purdue University, and Prof. Gregor Kemper from the Technical University of Munich. They developed a mathematical model, which they state proves that the system would work in real life.
The theoretical setup involves equipping an off-the-shelf quadcopter drone with a speaker, along with four microphones. The mics are set up in a three-dimensional arrangement, as opposed to all sitting flat relative to one another.
When the drone is subsequently flown through a room, its speaker emits short pulses of sound. For each pulse, the microphone array detects both the initial emission, and the subsequent echoes reflected back by the room's four walls.
The amount of time that elapses between the emission and any one echo is used to determine how far away the reflecting wall is … not unlike the manner in which bats use echolocation. And utilizing a mathematical method developed by the researchers – known as echo sorting – it's possible to determine which of the measured distances apply to which of the walls.
As a result, the system could continuously ascertain where the aircraft was relative to all of the walls within the room.
It is hoped that once developed further, the technology might be utilized to guide drones in dark, rainy/snowy or murky conditions where systems based on computer vision, lasers or infrared light may be ineffective. The setup could conceivably also find use in smartphone-based systems for guiding blind pedestrians, or in collision-avoidance systems for self-driving cars.
"The idea is to give the drone, or any other unmanned vehicle, the ability to navigate using sound," says Kemper. "Cars are already equipped with cameras. A new approach might be to include an acoustic sensor as well, to enhance the visual information already available and give a better picture of reality."
A paper on the research was recently published in the SIAM Journal on Applied Algebra and Geometry.