국립 부경대학교

Professor Lim Hae-kyun’s Team at Pukyong National University Developed Ultrasonic Dual Diagnosis System for Water Supply Pipes through Global Collaboration

- Simultaneous Measurement of Water Supply Pipe Corrosion and Water Quality... Published in the Nature Sister Journal <npj Clean Water>

Professor Lim Hae-kyun of the Biomedical Engineering Department at Pukyong National University and undergraduate student Seong Young-ho’s research team developed a system to simultaneously measure the corrosion of water supply pipes and water quality through global collaboration with Professor Lee O-jun of Catholic University and Dr. Hwan Ryul Jo from Flowserve Corporation, USA.

The paper titled ‘Internal pipe corrosion assessment method in water distribution system using ultrasound and convolutional neural networks ,‘ which contains the results of this research, was recently published in the Nature Sister Journal <npj Clean Water>(IF: 10.4, JCR 1.6% in Water Resources).

Iron oxide deposits resulting from pipe corrosion can contaminate water, leading to serious health issues such as gastrointestinal infections, skin problems, and lymph node complications, and if the corrosion weakens the pipe walls, the risk of leaks or ruptures increases, potentially resulting in higher repair costs and interruptions in water supply.

Accordingly, there has been a demand for an evaluation method that allows for non-destructive and continuous corrosion monitoring, enabling the early assessment of pipe conditions and appropriate maintenance to protect water quality and extend the lifespan of the pipes. Existing corrosion assessment methods often require damaging the pipes or interrupting the systems, leading to economic inefficiencies.

The research team developed a dual diagnosis system capable of non-destructive and continuous monitoring using ultrasound and artificial intelligence. They successfully generated high-resolution pipe thickness images using a high-frequency scanning acoustic microscope(SAM) to monitor the degree of pipe corrosion. At the same time, they analyzed ultrasonic signals in the pipes through convolutional neural networks(CNN) to measure the concentration of iron oxide in water.

In this study, pipes with a thickness reduction of 69~80 μm due to corrosion were measured using a high-frequency microscope, and all measurements showed an error margin within 5%. Additionally, the concentration of iron oxide in the pipes was classified using CNN, resulting in a high accuracy of up to 99%. It is expected that the dual diagnosis system capable of analyzing pipe corrosion and water quality will enable efficient and precise management of pipe infrastructure when applied in industrial settings.

Professor Lim Hae-kyun’s research team conducted this study through BK21 Four, the National Research Foundation of Korea’s Excellent Early-Career Researcher Support Project, the Regional Innovation Leading Research Center (RLRC), and the Excellent Researcher Exchange Support Project (BrainLink), and published their paper in <npj Clean Water> on July 13th.