In FaceOri, we needed to calibrate the system by putting the speakers against the microphone for a few seconds, which prevented this system from being widely used. To address this issue, I led a project developing a calibration-free distributed ultrasonic ranging module. This ongoing project aims to go beyond head motion tracking and build a module that could be integrated into sensor networks and support context-aware interaction leveraging multimodal data.

My Contributions

• Led this project and built the hardware prototype which enabled the FMCW-based ultrasonic ranging with a 100kHz sampling rate on nRF52840-DK using the tweeter and high sensitivity microphone.
• Enabled calibration-free ranging by leveraging Bluetooth for the time synchronization between two boards.
• Achieved 2.5 cm tracking accuracy in the range of 5 m.