This research aims to detect the leakage of pipeline by computerized on-line system using pressure analysis, as a determinant of the leakage in a pipe. At the first stage, the data is obtained from pressure changed at each location of the leakage and taken from the EPANET, a hydraulic modelling system, as simulated data. The simulation data consist of input data, in the form of pressure at each junction, and the output data, in the form of magnitude and location of leakage. Furthermore, the

more »

... is processed using one of the Artificial Intelligence methods, The Radial Basis Function Neural Network (RBF-NN), which has two phases: the learning and testing phases. The test results of the method of Radial Basis Function Neural Network are proven to be able to detect the magnitude and the location of leakage with the 98 % accurate prediction result of the whole pipeline system. The next step is creating pressure monitoring equipment on-line to replace the pressure data from the EPANET to the real data, thus the pressure at each junction can be monitored in real time. And by applying the method of RBF-NN, magnitude and location of leakage can be known.