skip to main content

Water distribution systems (WDSs) face a significant challenge in the form of pipe leaks. Pipe leaks can cause loss of a large amount of treated water, leading to pressure loss, increased energy costs, and contamination risks. Locating pipe leaks has been a constant challenge for water utilities and stakeholders due to the underground location of the pipes. Physical methods to detect leaks are expensive, intrusive, and heavily localized. Computational approaches provide an economical alternative to physical methods. Data-driven machine learning-based computational approaches have garnered growing interest in recent years to address the challenge of detecting pipe leaks in WDSs. While several studies have applied machine learning models for leak detection on single pipes and small test networks, their applicability to the real-world WDSs is unclear. Most of these studies simplify the leak characteristics and ignore modeling and measuring device uncertainties, which makes the scalability of their approaches questionable to real-world WDSs. Our study addresses this issue by devising four study cases that account for the realistic leak characteristics (multiple, multi-size, and randomly located leaks) and incorporating noise in the input data to account for the model- and measuring device- related uncertainties. A machine learning-based approach that uses simulated pressure as input to predict both location and size of leaks is proposed. Two different machine learning models: Multilayer Perceptron (MLP) and Convolutional Neural Network (CNN), are trained and tested for the four study cases, and their performances are compared. The precision and recall results for the L-Town network indicate good accuracies for both the models for all study cases, with CNN generally outperforming MLP.  more » « less
Award ID(s):
Author(s) / Creator(s):
Date Published:
Journal Name:
2nd International Joint Conference on Water Distribution Systems Analysis & Computing and Control in the Water Industry
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    Water‐hammer equations are used to describe transient flow in pipe networks. Uncertainty in model parameters, initial and boundary conditions, and location and strength of a possible leak renders deterministic predictions of this system untenable. When deployed in conjunction with pressure measurements, probabilistic solutions of the water‐hammer equations serve as a tool for detecting leaks in pipes. We use the method of distributions to obtain a probability density function (PDF) for pressure head, whose dynamics are described by the stochastic water‐hammer equations. This PDF provides a prior distribution for subsequent Bayesian data assimilation, in which data collected by pressure sensors are combined with this prior to obtain a posterior PDF of the leak location and size. We conduct a series of numerical experiments with uncertain initial velocity and measurement noise to ascertain the robustness and accuracy of the proposed approach. The results show the method's ability to identify the leak location and strength in a water transmission main.

    more » « less
  2. It is estimated that about 20% of treated drinking water is lost through distribution pipeline leakages in the United States. Pipeline leakage detection is a top priority for water utilities across the globe as leaks increase operational energy consumption and could also develop into potentially catastrophic water main breaks, if left unaddressed. Leakage detection is a laborious task often limited by the financial and human resources that utilities can afford. Many conventional leak detection techniques also only offer a snapshot indication of leakage presence. Furthermore, the reliability of many leakage detection techniques on plastic pipelines that are increasingly preferred for drinking water applications is questionable. As part of a smart water utility framework, this paper proposes and validates a hydraulic model-based technique for detecting and assessing the severity of leakages in buried water pipelines through monitoring of pressure from across the water distribution system (WDS). The envisioned smart water utility framework entails the capabilities to collect water consumption data from a limited number of WDS nodes and pressure data from a limited number of pressure monitoring stations placed across the WDS. A popular benchmark WDS is initially modified by inducing leakages through addition of orifice nodes. The leakage severity is controlled using emitter coefficients of the orifice nodes. WDS pressure data for various sets of demands is subsequently gathered from locations where pressure monitoring stations are to be placed in that modified distribution network. An evolutionary optimization algorithm is subsequently used to predict the emitter coefficients so as to determine the leakage severities based on the hydraulic dependency of the monitored pressure data on various sets of nodal demands. Artificial neural networks (ANNs) are employed to mimic the popular hydraulic solver EPANET 2.2 for high computational efficiency. The goals of this study are to: (1) validate the proof of concept of the proposed modeling approach for detecting and assessing the severity of leakages and (2) evaluate the sensitivity of the prediction accuracy to number of pressure monitoring stations and number of demand nodes at which consumption data is gathered and used. This study offers new value to prioritize pipes for rehabilitation by predicting leakages through a hydraulic model-based approach. 
    more » « less
  3. Pulmonary air leak is the most common complication of lung surgery, with air leaks that persist longer than 5 days representing a major source of post-surgery morbidity. Clinical management of air leaks is challenging due to limited methods to precisely locate and assess leaks. Here, we present a sound-guided methodology that enables rapid quantitative assessment and precise localization of air leaks by analyzing the distinct sounds generated as the air escapes through defective lung tissue. Air leaks often present after lung surgery due to loss of tissue integrity at or near a staple line. Accordingly, we investigated air leak sounds from a focal pleural defect in a rat model and from a staple line failure in a clinically relevant swine model to demonstrate the high sensitivity and translational potential of this approach. In rat and swine models of free-flowing air leak under positive pressure ventilation with intrapleural microphone 1 cm from the lung surface, we identified that: (a) pulmonary air leaks generate sounds that contain distinct harmonic series, (b) acoustic characteristics of air leak sounds can be used to classify leak severity, and (c) precise location of the air leak can be determined with high resolution (within 1 cm) by mapping the sound loudness level across the lung surface. Our findings suggest that sound-guided assessment and localization of pulmonary air leaks could serve as a diagnostic tool to inform air leak detection and treatment strategies during video-assisted thoracoscopic surgery (VATS) or thoracotomy procedures. 
    more » « less
  4. Zelinski, Michael E. ; Taha, Tarek M. ; Howe, Jonathan (Ed.)
    Image classification forms an important class of problems in machine learning and is widely used in many realworld applications, such as medicine, ecology, astronomy, and defense. Convolutional neural networks (CNNs) are machine learning techniques designed for inputs with grid structures, e.g., images, whose features are spatially correlated. As such, CNNs have been demonstrated to be highly effective approaches for many image classification problems and have consistently outperformed other approaches in many image classification and object detection competitions. A particular challenge involved in using machine learning for classifying images is measurement data loss in the form of missing pixels, which occurs in settings where scene occlusions are present or where the photodetectors in the imaging system are partially damaged. In such cases, the performance of CNN models tends to deteriorate or becomes unreliable even when the perturbations to the input image are small. In this work, we investigate techniques for improving the performance of CNN models for image classification with missing data. In particular, we explore training on a variety of data alterations that mimic data loss for producing more robust classifiers. By optimizing the categorical cross-entropy loss function, we demonstrate through numerical experiments on the MNIST dataset that training with these synthetic alterations can enhance the classification accuracy of our CNN models. 
    more » « less
  5. null (Ed.)
    The DeepLearningEpilepsyDetectionChallenge: design, implementation, andtestofanewcrowd-sourced AIchallengeecosystem Isabell Kiral*, Subhrajit Roy*, Todd Mummert*, Alan Braz*, Jason Tsay, Jianbin Tang, Umar Asif, Thomas Schaffter, Eren Mehmet, The IBM Epilepsy Consortium◊ , Joseph Picone, Iyad Obeid, Bruno De Assis Marques, Stefan Maetschke, Rania Khalaf†, Michal Rosen-Zvi† , Gustavo Stolovitzky† , Mahtab Mirmomeni† , Stefan Harrer† * These authors contributed equally to this work † Corresponding authors:,,,, ◊ Members of the IBM Epilepsy Consortium are listed in the Acknowledgements section J. Picone and I. Obeid are with Temple University, USA. T. Schaffter is with Sage Bionetworks, USA. E. Mehmet is with the University of Illinois at Urbana-Champaign, USA. All other authors are with IBM Research in USA, Israel and Australia. Introduction This decade has seen an ever-growing number of scientific fields benefitting from the advances in machine learning technology and tooling. More recently, this trend reached the medical domain, with applications reaching from cancer diagnosis [1] to the development of brain-machine-interfaces [2]. While Kaggle has pioneered the crowd-sourcing of machine learning challenges to incentivise data scientists from around the world to advance algorithm and model design, the increasing complexity of problem statements demands of participants to be expert data scientists, deeply knowledgeable in at least one other scientific domain, and competent software engineers with access to large compute resources. People who match this description are few and far between, unfortunately leading to a shrinking pool of possible participants and a loss of experts dedicating their time to solving important problems. Participation is even further restricted in the context of any challenge run on confidential use cases or with sensitive data. Recently, we designed and ran a deep learning challenge to crowd-source the development of an automated labelling system for brain recordings, aiming to advance epilepsy research. A focus of this challenge, run internally in IBM, was the development of a platform that lowers the barrier of entry and therefore mitigates the risk of excluding interested parties from participating. The challenge: enabling wide participation With the goal to run a challenge that mobilises the largest possible pool of participants from IBM (global), we designed a use case around previous work in epileptic seizure prediction [3]. In this “Deep Learning Epilepsy Detection Challenge”, participants were asked to develop an automatic labelling system to reduce the time a clinician would need to diagnose patients with epilepsy. Labelled training and blind validation data for the challenge were generously provided by Temple University Hospital (TUH) [4]. TUH also devised a novel scoring metric for the detection of seizures that was used as basis for algorithm evaluation [5]. In order to provide an experience with a low barrier of entry, we designed a generalisable challenge platform under the following principles: 1. No participant should need to have in-depth knowledge of the specific domain. (i.e. no participant should need to be a neuroscientist or epileptologist.) 2. No participant should need to be an expert data scientist. 3. No participant should need more than basic programming knowledge. (i.e. no participant should need to learn how to process fringe data formats and stream data efficiently.) 4. No participant should need to provide their own computing resources. In addition to the above, our platform should further • guide participants through the entire process from sign-up to model submission, • facilitate collaboration, and • provide instant feedback to the participants through data visualisation and intermediate online leaderboards. The platform The architecture of the platform that was designed and developed is shown in Figure 1. The entire system consists of a number of interacting components. (1) A web portal serves as the entry point to challenge participation, providing challenge information, such as timelines and challenge rules, and scientific background. The portal also facilitated the formation of teams and provided participants with an intermediate leaderboard of submitted results and a final leaderboard at the end of the challenge. (2) IBM Watson Studio [6] is the umbrella term for a number of services offered by IBM. Upon creation of a user account through the web portal, an IBM Watson Studio account was automatically created for each participant that allowed users access to IBM's Data Science Experience (DSX), the analytics engine Watson Machine Learning (WML), and IBM's Cloud Object Storage (COS) [7], all of which will be described in more detail in further sections. (3) The user interface and starter kit were hosted on IBM's Data Science Experience platform (DSX) and formed the main component for designing and testing models during the challenge. DSX allows for real-time collaboration on shared notebooks between team members. A starter kit in the form of a Python notebook, supporting the popular deep learning libraries TensorFLow [8] and PyTorch [9], was provided to all teams to guide them through the challenge process. Upon instantiation, the starter kit loaded necessary python libraries and custom functions for the invisible integration with COS and WML. In dedicated spots in the notebook, participants could write custom pre-processing code, machine learning models, and post-processing algorithms. The starter kit provided instant feedback about participants' custom routines through data visualisations. Using the notebook only, teams were able to run the code on WML, making use of a compute cluster of IBM's resources. The starter kit also enabled submission of the final code to a data storage to which only the challenge team had access. (4) Watson Machine Learning provided access to shared compute resources (GPUs). Code was bundled up automatically in the starter kit and deployed to and run on WML. WML in turn had access to shared storage from which it requested recorded data and to which it stored the participant's code and trained models. (5) IBM's Cloud Object Storage held the data for this challenge. Using the starter kit, participants could investigate their results as well as data samples in order to better design custom algorithms. (6) Utility Functions were loaded into the starter kit at instantiation. This set of functions included code to pre-process data into a more common format, to optimise streaming through the use of the NutsFlow and NutsML libraries [10], and to provide seamless access to the all IBM services used. Not captured in the diagram is the final code evaluation, which was conducted in an automated way as soon as code was submitted though the starter kit, minimising the burden on the challenge organising team. Figure 1: High-level architecture of the challenge platform Measuring success The competitive phase of the "Deep Learning Epilepsy Detection Challenge" ran for 6 months. Twenty-five teams, with a total number of 87 scientists and software engineers from 14 global locations participated. All participants made use of the starter kit we provided and ran algorithms on IBM's infrastructure WML. Seven teams persisted until the end of the challenge and submitted final solutions. The best performing solutions reached seizure detection performances which allow to reduce hundred-fold the time eliptologists need to annotate continuous EEG recordings. Thus, we expect the developed algorithms to aid in the diagnosis of epilepsy by significantly shortening manual labelling time. Detailed results are currently in preparation for publication. Equally important to solving the scientific challenge, however, was to understand whether we managed to encourage participation from non-expert data scientists. Figure 2: Primary occupation as reported by challenge participants Out of the 40 participants for whom we have occupational information, 23 reported Data Science or AI as their main job description, 11 reported being a Software Engineer, and 2 people had expertise in Neuroscience. Figure 2 shows that participants had a variety of specialisations, including some that are in no way related to data science, software engineering, or neuroscience. No participant had deep knowledge and experience in data science, software engineering and neuroscience. Conclusion Given the growing complexity of data science problems and increasing dataset sizes, in order to solve these problems, it is imperative to enable collaboration between people with differences in expertise with a focus on inclusiveness and having a low barrier of entry. We designed, implemented, and tested a challenge platform to address exactly this. Using our platform, we ran a deep-learning challenge for epileptic seizure detection. 87 IBM employees from several business units including but not limited to IBM Research with a variety of skills, including sales and design, participated in this highly technical challenge. 
    more » « less