Smart Water Meter

Citation Author(s):
Vincenzo
Gallo
University of Salerno
Salvatore
Dello Iacono
University of Salerno
Marco
Carratù
University of Salerno
Giuseppe
Di Leo
University of Salerno
Consolatina
Liguori
University of Salerno
Antonio
Pietrosanto
University of Salerno
Submitted by:
Vincenzo Gallo
Last updated:
Tue, 12/13/2022 - 11:14
DOI:
10.21227/395m-cy10
Data Format:
License:
458 Views
Categories:
Keywords:
0
0 ratings - Please login to submit your rating.

Abstract 

Water leakage problems increased over the last few years, and innovative tools and techniques appeared to solve this widespread problem. The still unresolved problem concerns the identification of water leaks at the nearest point; at the household level, the most common and inexpensive devices are still mechanical meters, which cannot detect leaks. While the issue is not important for water service providers since consumption is charged to the user, the resolution is of crucial importance due to the increasingly relevant concern of saving natural resources. A previous proposal exploited classical vision algorithms on mechanical water meters but required a controlled environment, like illumination and target positioning, and time-consuming calibration procedures. Machine learning approaches enabled image processing techniques also in non-controlled environments, overcoming the classical methods but introducing new challenges like power consumption.I n this research, a water leakage detector at the household level based on machine learning is presented. Using a combination of Convolutional Neural Networks (CNNs) and Long Short-Term Memory Networks (LSTM) with an under-sampling technique, it is possible to process the images of a mechanical water meter dial and identify the Period With Null Consumption (PWNC) or the consumption class.

Instructions: 

The dataset consists of several images categorized by applied water flow (L/h).