Deep Learning for Acoustic-based Pipe Leak Detection

Current leak detection methods for water pipelines, such as fixed acoustic devices and Pipeline Inspection Gadgets (PIGs), require shutting off the water supply and are often impractical for field inspections. These methods demand extensive workforce and time and often fail to detect smaller leaks effectively, leading to significant water loss, increased operational costs, and potential infrastructure damage. XK Innovate Inc., a leader in water pipeline technology based in Toronto, Ontario, will collaborate with UBC Okanagan to develop a groundbreaking in-service water pipe leak detection system. This project introduces the HZ1 Free-Swimming Device, advanced deep learning models, and an intuitive Graphical User Interface (GUI) to detect leaks without disrupting the water supply. This innovative solution aims to reduce workforce requirements, operational costs, and water loss by enhancing accuracy and efficiency while fostering Canadian economic growth and public welfare. The project aligns with Canada’s innovation and workforce development policies and supports sustainable development by minimizing environmental impact and resource consumption. Additionally, it offers an invaluable opportunity for an intern to gain practical experience in data analytics, systems engineering, and environmental technology, equipping them with competitive skills for the job market.

Faculty Supervisor:

Zheng Liu

Student:

Partner:

XK Innovate, Inc.

Discipline:

Engineering

Sector:

Professional, scientific and technical services

University:

The University of British Columbia - Okanagan

Program: