Ultralow-power standoff acoustic leak detection

Michael P. Hasselbeck · 2025

Read Ultralow-power standoff acoustic leak detection by Michael P. Hasselbeck (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-00348.txt.

Introduction

Michael Hasselbeck describes a listening device that spots leaks from pressurized plumbing without touching the pipes. The design borrows ideas from glass-break and smoke detectors, raising an alert when leak sounds appear. In tests, a simulated water leak at a slow flow was detected reliably from more than ten meters away, even through walls, doors, floors, and ceilings.

The unit is built as a battery-powered wireless node that processes sound on the spot, sending only a few status bytes when asked. Power draw stays extremely low, in the tens to hundreds of microwatts, depending on noise and response speed. It listens above the range of human speech, so it cannot eavesdrop. Though developed for water, the concept also fits gas leaks.

Catalog record

CREATORMichael P. Hasselbeck
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-00348

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Ultralow-power standoff acoustic leak detection Michael P. Hasselbeck Abstract An automated, standoff acoustic leak detection scheme has been designed, built, and tested. It merges the principles of glass breakage and smoke detection to alert for the presence of leaks emanating from pressurized plumbing. A simulated water leak flowing at 0.15 l/min has been reliably detected at a standoff distance of more than 10 m. The device is also effective at identifying the presence of leaks located behind surfaces such as walls, doors, floors, and ceilings. The anticipated application is as an autonomous, battery-powered, remote wireless node. All signal processing and analysis takes place on the edge with no need to stream audio data to the cloud. Sensor status is conveyed on-demand with only a few bytes of information, requiring minimal bandwidth. Power consumption is the range of 20–200 μ \mu W, depending on the amount of environmental noise and desired sensor latency. To attain optimum sensitivity and reliability, the hardware operates at acoustic frequencies well above the range of human conversations, making eavesdropping impossible. Development has been done with water escaping from pressurized plumbing, but the sensor concept can be used effectively to detect gas leaks. I Introduction There are two methods currently available for autonomous water leak detection: contact moisture sensors [ 1 ] and smart flow meters [ 2 ] . They work well and can provide protection in many applications. Identifying the onset of very small leaks, however, can be difficult especially in situations where installing existing products may not be easy or convenient. This paper describes an automated, inexpensive, battery-powered, smart acoustic leak detector concept that addresses these problems and can augment the capabilities of a building monitoring system. When pressurized water escapes from a seam, crack, or loose fitting on plumbing, broadband high frequency acoustics are emitted. These signals can travel an extended distance ( > 10 m) in free-space and be detected by a remote sensor [ 3 , 4 ] . This suggests a standoff method for water leak detection that is substantially different compared to conventional contact moisture sensors and various types of smart flow meters installed on or inside of plumbing lines. Using acoustics to detect leaks is a well established approach and a variety of commercial devices are available. These sensors are generally sophisticated, expensive, and require a trained technician to operate [ 5 , 6 ] . It is desirable to completely automate the leak detection process with a smart device that can be deployed as a remote sensor node on a wireless network. The acoustic leak sensor combines the principles of glass breakage and smoke detectors. Glass breakage occurs on a timescale of a few seconds or less, requiring continuous data acquisition at rates approaching 1 kHz or more [ 7 ] . Almost all pressurized plumbing leaks are persistent, which allows for sub-Hz sampling rates and exceptional battery life. In a typical application, two AAA alkaline batteries should power the device for multiple years. Glass breakage has the advantage of being generally louder than the ambient environment so that detector sensitivity is not an issue. Small water leaks, especially when separated from the sensor by acoustic barriers such as walls, are likely to be at or below the environmental background. Much higher sensitivity and noise mitigation are crucial. This can be addressed by active (analog and digital) and passive (mechanical) filtering along with judicious use of statistical analysis. The acoustic approach to water leak detection described here is not capable of identifying all water issues. It will not respond to leaking roofs or melting ice, for example. The nature of the pressurized orifice is also important. A pinhole leak will produce a Laminar flow jet that emits a much weaker acoustic signal – perhaps undetectable – compared to a jagged crack or loose fitting at the same backing pressure. Acoustic sensing can be used to complement standard leak detection techniques and will be especially beneficial when monitoring large areas. Anticipated use cases are in smart homes and buildings, security systems, irrigation, dental and medical offices, critical plumbing installations, insurance loss mitigation, and water conservation. Although the design been developed for water leaks, it is expected to be useful in other applications that involve pressurized gases such as air, refrigerants, or steam. It is important to emphasize that there is no attempt to make quantitative measurements of flow. The sensor is listening for a persistent acoustic signal that indicates the possible presence of a leak. The leak rate is irrelevant; we are only interested in whether or not one is present. It targets the smallest leaks, below flow rates where utility meters may not respond. The acoustic leak detector is agnostic to the host hardware and will work with any controller having an I 2 C interface and 3V3 power bus. Best performance is attained in an environment with a stable, relatively quiet acoustic background. Human activity will often produce noise that will impair the sensor’s ability to hear leaks. The formation of damp spots, puddles, and similar indicators of water escaping, however, are usually obvious and quickly noticed. Highest reliability for an acoustic sensor will occur in conditions where people are not present. II Design The sensor is fabricated on a double-sided PCB using low cost, commercial electronic components (see Fig. 1 ). It is designed to work with a host controller as part of a low power wireless system with a well-filtered 1.8 to 3.6 Vdc power source. All sensor training, signal processing, and analysis occurs on the device; no audio is ever sent to the cloud. Two-way controller-sensor communication takes places via I 2 C. Firmware, a system flow chart, and design files are available in the GitHub repository [ 8 ] . Fig. 1: The sensor PCB measures 48 mm x 33 mm. A surface mount MEMS microphone is enclosed by a rectangular chamber to form a Helmholtz resonator (far left). The 7-pin connection header is used for programming the MCU, I 2 C communication, and 3V3 dc power. II-A Frequency Filtering It is essential to separate weak leak signals from the ambient background to prevent false triggers and reduce the likelihood of false alarms. Leaks are sufficiently broadband that they produce discernible acoustic energy at frequencies above 6 kHz; these are the frequencies targeted by the sensor. The lower frequencies are attenuated by a combination of mechanical, analog electronic, and digital filtering. Fig. 2: Design diagram of the Helmholtz resonator. A partially open rectangular chamber is placed over a MEMS microphone (red lines). The resonator wall thickness t is not shown for clarity. Mechanical filtering is implemented with a miniature Helmholtz resonator that is formed by a rectangular plastic chamber glued above a surface-mount MEMS microphone (Knowles SPU0410). Acoustic energy is admitted through a circular hole of radius a . Referring to Fig. 2 , the resonator center frequency f_{0} is given by [ 9 ] : f_{0}=\frac{va}{2}\sqrt{\frac{1}{\pi Vt^{\prime}}} (1) where v = 343 m/s is the speed of sound and V=LWH is the partially enclosed air volume above the microphone surface. The rectangular footprint of the microphone is L = 3.9 mm and W = 3.2 mm; H is the interior height

Read the complete text