Developer creates ai-assisted tool to pinpoint indoor sound sources within 40 centimeters

Software developer Simon Willison recently unveiled Beep Locator, a sound localization system built using the AI programming tool Claude Code. The system is capable of narrowing down the origin of a sound to an approximate radius of 40 centimeters. Willison developed the tool after encountering intermittent, short electronic beeps emanating from unopened boxes in his storage room, which were difficult to locate by ear alone. The application utilizes multiple devices—five phones, tablets, and computers—to collectively capture audio via web-based microphone access. Technically, Beep Locator functions by analyzing the time differential of a sound arriving at various devices. To compensate for initial inaccuracies in automated positioning, the system incorporates manual map dragging and device placement annotation. Furthermore, it employs Fast Fourier Transform analysis to isolate the target tone from ambient noise, synchronizing recordings across all endpoints using a short sound emitted by one device. The project code has been made publicly available on GitHub.