MetaSonic: Advancing Robot Localization With Directional Embedded Acoustic Signals

Junling Wang, Zhenlin An, Yi Guo

Abstract: Indoor positioning in environments where GPS cannot be used is a fundamental technology for robot navigation and human-robot interaction. However, existing vision-based localization systems cannot work in low-visibility environments, and existing wireless or acoustic localization systems require specific transceivers, making them expensive and power-intensive — particularly challenging for microrobots. This letter proposes a new metasurface-assisted ultrasound positioning system. The key idea is to use a low-cost passive acoustic metasurface to transfer any speaker into a directional sound source, with the acoustic spectrum varying based on direction. This allows any microrobot with a simple, low-cost microphone to capture such modified sound to identify the direction of the sound source. We develop a lightweight convolutional neural network-based localization algorithm that can be efficiently deployed on low-power microcontrollers. We evaluate our system in a large complex office. It can achieve a direction estimation accuracy of 7.26$^\circ$, improving by 42.2% compared to systems without the metasurface and matching the performance of a 4-microphone array, with a localization accuracy of 0.35 m.

External IDs:dblp:journals/ral/WangAG25