John Mashford

PIRAT—A System for Quantitative Sewer Pipe Assessment

Sewers are aging, expensive assets that attract public attention only when they fail. Sewer operators are under increasing pressure to minimise their maintenance costs, while preventing sewer failures. Inspection can give early warning of failures and allow economical repair under noncrisis conditions. Current inspection techniques are subjective and detect only gross defects reliably. They cannot provide the data needed to confidently plan long-term maintenance. This paper describes PIRAT, a quantitative technique for sewer inspection.PIRAT measures the internal geometry of the sewer and then analyses these data to detect, classify, and rate defects automatically using artificial intelligence techniques. We describe the measuring system and present and discuss geometry results for different types of sewers. The defect analysis techniques are outlined and a sample defect report presented. PIRAT’s defect reports are compared with reports from the conventional technique and the discrepancies discussed. We relate PIRAT to other work in sewer robotics.

Superpixels via pseudo-Boolean optimization

We propose an algorithm for creating superpixels. The major step in our algorithm is simply minimizing two pseudo-Boolean functions. The processing time of our algorithm on images of moderate size is only half a second. Experiments on a benchmark dataset show that our method produces superpixels of comparable quality with existing algorithms. Last but not least, the speed of our algorithm is independent of the number of superpixels, which is usually the bottle-neck for the traditional algorithms of superpixel creation.

An Approach to Leak Detection in Pipe Networks Using Analysis of Monitored Pressure Values by Support Vector Machine

This paper presents a method of mining the data obtained by a collection of pressure sensors monitoring a pipe network to obtain information about the location and size of leaks in the network. This inverse engineering problem is effected by support vector machines (SVMs) which act as pattern recognisers. In this study the SVMs are trained and tested on data obtained from the EPANET hydraulic modelling system. Performance assessment of the SVM showed that leak size and location are both predicted with a reasonable degree of accuracy. The information obtained from this SVM analysis would be invaluable to water authorities in overcoming the ongoing problem of leak detection.

Prediction of Sewer Condition Grade Using Support Vector Machines

Assessing the condition of sewer networks is an important asset management approach. However, because of high inspection costs and limited budget, only a small proportion of sewer systems may be inspected. Tools are therefore required to help target inspection efforts and to extract maximum value from the condition data collected. Owing to the difficulty in modeling the complexities of sewer condition deterioration, there has been interest in the application of artificial intelligence-based techniques such as artificial neural networks to develop models that can infer an unknown structural condition based on data from sewers that have been inspected. To this end, this study investigates the use of support vector machine (SVM) models to predict the condition of sewers. The results of model testing showed that the SVM achieves good predictive performance. With access to a representative set of training data, the SVM modeling approach can therefore be used to allocate a condition grade to sewer assets with reasonable confidence and thus identify high risk sewer assets for subsequent inspection.

A neural network image classification system for automatic inspection

The development of a neural network image classification system for range image automatic inspection is described. The classifier operates on regions of interest (ROIs) which have been identified in the image through segmentation and connected component labelling. The classification of ROIs is carried out by a Bayesian decision tree of feedforward neural networks operating on features derived from the region and image model. The feature set includes Fourier wedge-ring samples and image histogram moments.

Risk-based prioritization and its application to inspection of valves in the water sector

Isolation valves facilitate the effective operation and maintenance of water supply networks, but their sheer number presents a significant asset management challenge. If left unmanaged, valve reliability issues can become widespread. Inspections provide a means of increasing reliability, but a survey of industry practices indicated that some utilities did not have such a program in place. To improve asset management and reduce business risk exposure, such utilities need an effective means of commencing inspection programs. From a theoretical perspective, risk concepts provide a means of optimizing maintenance effort. However, in the face of poor data on reliability or condition, pragmatic approaches to risk-based prioritization are needed. One such approach, risk indexing, is considered in this paper. Background on the research is presented, including the application of risk-based inspection concepts within the water sector. The development of a risk indexing scheme is then investigated, drawing on two industry workshops in which the analytical hierarchy process was used to set relative weights. It is concluded that risk indexing provides the basis for a rational prioritization process in the absence of data on valve reliability or condition.

LEAK DETECTION IN SIMULATED WATER PIPE NETWORKS USING SVM

The detection and location of leaks in water pipe networks is a significant problem, which would benefit from more effective solutions. The information about the presence and location of leaks in a pipe network could be contained in the distribution of pressure or flow values at various points in the network; however, the information is encoded in such a way that its extraction is a complex inverse engineering problem. Such problems can be solved effectively through the use of pattern recognition techniques such as artificial neural networks (ANNs) or support vector machines (SVMs). This article presents a method of using SVM analysis to interpret the data obtained by a collection of pressure sensors or flow-measuring devices monitoring a pipe network in order to obtain information about the location and size of leaks in the network.

Pixel-based colour image segmentation using support vector machine for automatic pipe inspection

This paper presents a new approach to image segmentation of colour images for automatic pipe inspection. Pixel-based segmentation of colour images is carried out by a support vector machine (SVM) labelling pixels on the basis of local features. Segmentation can be effected by this pixel labelling together with connected component labelling. The method has been tested using RGB, HSB, Gabor, local window and HS feature sets and is seen to work best with the HSB feature set.

Development of a Fuzzy Risk Ranking Model for Prioritizing Manhole Inspection

Manholes are designed to provide access points to underground sewer networks for inspection and maintenance. Manhole collapses, although rare, can result in severe consequences and have a significant effect on the performance of the sewer. This paper presents a case study on the development of a risk ranking model using fuzzy set theory and the analytical hierarchy process for individual manholes of sewer networks. The fuzzy risk ranking model (FRM) considered both the likelihood and consequence of collapse. The performance of the FRM was validated against 10 manholes with known poor condition. The results were also compared against a previously developed risk ranking scheme, with regards to consistency and repeatability of the relative ranking of assets. The results suggested that the FRM may provide better performance, although only limited data were available for validation. The process adopted in constructing the scheme is considered to be systematic and auditable.

An Approach to Pipe Image Interpretation Based Condition Assessment for Automatic Pipe Inspection

Condition assessment forms an important part of the asset management of buried pipelines. This is carried out through the use of inspection systems which usually consist of an image acquisition device attached to a mobile robotic platform. Complete or partial automation of image interpretation could increase the efficiency and objectivity of pipe inspection. A key component of an automatic pipe inspection system is the segmentation module. This paper describes an approach to automatic pipe inspection using pixel-based segmentation of colour images by support vector machine (SVM) coupled with morphological analysis of the principal component of the segmented image. The morphological analysis allows the principal component of the segmented image to be decomposed into the pipe flow lines region, the pipe joints, and adjoining defects. A simple approach to detecting pipe connections using fuzzy membership functions relating to defect size and location is also described.