Water-leak detection will save millions
A research consortium in the UK is working on a project which promises to save the water industry - and its customers - millions of pounds (euro) by developing a computer system to predict and detect water leaks in underground pipelines. In a three-year, £1.2 million ( 2 million euro) Department of Trade and Industry initiative, the Engineering and Physical Sciences Research Council (EPSRC) has awarded grants to four universities - Bradford, East Anglia, Brunel and Imperial College - to form a consortium to tackle the problem. Bradford's share of the research funding is £310,306. Professor Imad Torsun, Department of Computing, Professor Andrew Day, Department of Mechanical and Medical Engineering, and Dr Alastair Wood, Mathematics lead its team. Professor Torsun is developing a neural network - computing software - that will predict and detect leaks. The team will also develop a knowledge-based system based on the neural network, which will generate remedies such as turning off a particular valve on the affected pipeline. To create a prototype software system, Professor Torsun will use data collected by sensors already in place on sections of Yorkshire Water pipeline at Keighley which log measurements of the flow, pressure and temperature of water every 15 minutes. Dr Wood will measure the pre-processing and reliability of the data. Imperial College and Brunel University are also working on mathematical models, which will help to measure and verify the quality and reliability of the Bradford data. Professor Day is developing an innovative, low-cost and reliable sensor for eventual use throughout the pipeline network. Once the neural network has been developed, the Bradford team should be able to locate leaks to within half-a-kilometer on a section of pipeline. The team works closely with colleagues at East Anglia who use acoustic water-contact sensors to pinpoint the leak to within a few meters. "We tell them roughly where the leak is and they detect it precisely," said Professor Torsun. "Being able to identify leaks automatically will save the water companies millions of pounds (euro). This is an exciting project using teamwork within Bradford and collaboration with other universities and industry. But it is not just blue-sky research. It's at the cutting edge of technology and will have immediate and very effective applications to an extremely vital commodity - water."
Countries
United Kingdom