Sound location technology is often designed around the human ear, but why not when it is a bat? This is clearly good? Virginia Tech researchers have certainly asked this question. They did Advanced A sound location system that accompanies a bat-like ear design with a deep neural network for determining sound within half a degree – a pair of human ears accurate to nine degrees, even the latest technology stops at 7.5 degrees.
The system shakes the outer ear to create a Doppler shift signature relative to the sound source. Since the patterns are easily complex for Decker, the team trained the neural network to direct the source of each received echo. And unlike human-inspired systems, it only requires one receiver and a single frequency.
Much work remains to be done before it becomes practical, but scientists are already seeing their bat-based technology help robots navigate “complex” outdoor spaces with little work. You get to see Farm robot Crops with high precision, Biodiversity Monitor Indomitable precision and military robots that cross difficult terrain with little input. If nothing else, it is a reminder that the most effective biomimetry is not necessarily based on our own biology.
All products offered by Engadget are selected by our editorial team, different from our parent company. Some of our stories include approved links. If you purchase something through one of these links, we can earn an approved commission.