Faulty Data Research Articles

Logistics 4.0 and supply chain 4.0 in the automotive industry

The Industry 4.0 technologies give real benefits to the plants where they are implemented such as performance improvement, costs and delays reductions. The automotive plants are among the factories that could implement most of these technologies. Industry 4.0 can affect all of the processes which happen in the automotive industry manufacturing plants – engineering, production, logistics, management, etc. Analysis of all of the processes in one article is too complex task and due to this the paper is focused only on logistics and supply chain. The aim of this article is to investigate the potential possibilities for implementation of Logistics 4.0 and Supply Chain 4.0 in the automotive industry plants, in particular in Bulgaria. Specially designed survey was used as investigation method. It consists of two groups with proposals and assumptions. The survey was conducted by 12 automotive industry logistics experts who are on 6 different logistics positions. The results show that the production visibility (assembly lines connected to internet) and blockchain technologies have the biggest values to the logistics and supply chain. But the cybersecurity and faulty data are the main challenges in front of them.

RFI Artefacts Detection in Sentinel-1 Level-1 SLC Data Based On Image Processing Techniques.

Interferometric Synthetic Aperture Radar (InSAR) data are often contaminated by Radio-Frequency Interference (RFI) artefacts that make processing them more challenging. Therefore, easy to implement techniques for artefacts recognition have the potential to support the automatic Permanent Scatterers InSAR (PSInSAR) processing workflow during which faulty input data can lead to misinterpretation of the final outcomes. To address this issue, an efficient methodology was developed to mark images with RFI artefacts and as a consequence remove them from the stack of Synthetic Aperture Radar (SAR) images required in the PSInSAR processing workflow to calculate the ground displacements. Techniques presented in this paper for the purpose of RFI detection are based on image processing methods with the use of feature extraction involving pixel convolution, thresholding and nearest neighbor structure filtering. As the reference classifier, a convolutional neural network was used.

Data‐driven nonlinear chemical process fault diagnosis based on hierarchical representation learning

AbstractRepresentation extraction is crucial in data‐driven process monitoring, and deep neural network (DNN) is an efficient tool for extracting representations from considerable process data. This study proposes a hierarchical representation learning (HRL) method that integrates the deep belief neural (DBN) network and support vector data description (SVDD) for efficient nonlinear chemical process fault diagnosis. First, hierarchical representations containing meaningful process information are generated through a DBN network by utilizing generally massive normal operating process data. Second, an SVDD‐based decision‐making system is constructed using generally small‐sized faulty data. Three experimental studies are then conducted. A comparison of results with those of several state‐of‐the‐art methods reveal the suitability of the HRL method for process monitoring due to its two main advantages. First, DNN has a superior representative ability and generates representations with richer process information than conventional data‐driven methods. Second, the HRL method utilizes available process data and is suitable for practical conditions in which considerable normal operating data but limited small‐sized faulty data are available.

Child porn users & risk for engaging in contact offenses: faulty data minimizes offender’s risk & puts more children at risk for sexual abuse

Child porn users & risk for engaging in contact offenses: faulty data minimizes offender’s risk & puts more children at risk for sexual abuse

Unsupervised deep representation learning for motor fault diagnosis by mutual information maximization

Data-driven deep learning technology has gained many achievements in the field of motor fault diagnosis and prognostics. However, the application objects of those previous studies are commonly limited to the faulty data sharing the similar distribution under unvarying stable working condition. Unfortunately, this limitation is nearly invalid in the real-world scenario, where the working condition is complicated and changes invariably, resulting in the unfavourable situation that the deep representation learning methods of the previous studies always fail in extracting the effective representations for fault diagnosis in real applications. To tackle this issue, inspired by f-divergence estimation, this work takes a different route and proposes an unsupervised deep representation learning approach, named Deep Mutual Information Maximization (DMIM), using variational divergence estimation approach to maximize mutual information (MI) between the input and output of a deep neural network. Meanwhile the representation distribution is automatically tuned by matching to a prior distribution with the same philosophy of Variational Autoencoder. Opposite to previous works which learn representations basically with supervised feedback regulation or unsupervised reconstruction, the proposed unsupervised MI maximization framework aims to make representational characteristics like independence play a bigger role to capture the most unique representations. To verify the effectiveness of our proposal, faulty motor data from the motor tests under European driving cycle for simulating the real working scenario, are collected for validation. It turns out that DMIM outperforms many popular unsupervised and fully-supervised learning methods. It opens new avenues for unsupervised learning of representations for motor fault diagnosis.

Comparison of Machine Learning Techniques for Software Quality Prediction

Software quality prediction is one the most challenging tasks in the development and maintenance of software. Machine learning (ML) is widely being incorporated for the prediction of the quality of a final product in the early development stages of the software development life cycle (SDLC). An ML prediction model uses software metrics and faulty data from previous projects to detect high-risk modules for future projects, so that the testing efforts can be targeted to those specific ‘risky' modules. Hence, ML-based predictors contribute to the detection of development anomalies early and inexpensively and ensure the timely delivery of a successful, failure-free and supreme quality software product within budget. This article has a comparison of 30 software quality prediction models (5 technique * 6 dataset) built on five ML techniques: artificial neural network (ANN); support vector machine (SVMs); Decision Tree (DTs); k-Nearest Neighbor (KNN); and Naïve Bayes Classifiers (NBC), using six datasets: CM1, KC1, KC2, PC1, JM1, and a combined one. These models exploit the predictive power of static code metrics, McCabe complexity metrics, for quality prediction. All thirty predictors are compared using a receiver operator curve (ROC), area under the curve (AUC), and accuracy as performance evaluation criteria. The results show that the ANN technique for software quality prediction is promising for accurate quality prediction irrespective of the dataset used.

SF6 fault decomposition feature component extraction and triangle fault diagnosis method

How to use SF 6 decomposition feature component information to judge the form and degree of gas-insulated equipment (GIE) field discharge and overthermal faults quickly is a problem that remains unresolved. Based on the existing experimental data on SF 6 typical fault decomposition, this study considers the SF 6 decomposition mechanism under typical faults and uses the maximum correlation minimum redundancy criterion to filter out three decomposition feature components characterizing GIE typical fault attributes: SOF 2 +SO 2 , CF 4 , and SO 2 F 2 . The weight of extracted feature components is optimized by the area equivalence principle, and the triangle fault diagnosis method of the SF 6 decomposition component that is suitable for the rapid diagnosis of a GIE field is constructed. The diagnosis method is comprehensively tested using faulty data in different conditions, and the comprehensive recognition rate of lab tests reaches 96.2%. Results show that the constructed triangle fault diagnosis method of the SF 6 decomposition component can diagnose the internal fault nature of a GIE and identify the types of insulation defects that induce partial discharge faults. Moreover, the constructed method in this research is simple, effective, and suitable for field maintenance and online intelligent monitoring of GIE.

Towards Generating Realistic and High Coverage Test Data for Constraint-Based Fault Injection

Generating faulty data is a key issue in fault injection. The faulty data include not only the ones of extreme values or bad formats, but also the ones which are logically unreasonable. Constraint-based fault injection which negates interface constraints to solve faulty data is effective for logically unreasonable data generation. However, the existing constraint-based approaches just solve brand new data for testing. Such brand new data may easily violate some hidden environment constraints on the test inputs and hence be nonrealistic. Besides, there can be different strategies to negate a constraint in order to derive the constraint-unsatisfied faulty data. What are the possible negation strategies and which strategies are better for high coverage fault injection are still unclear. To these ends, this paper presents a new constraint-based fault injection approach. The approach introduces 10 different strategies for constraint negation and relaxes constraint variables to generate faulty data instead of solving brand new data for fault injection. It can produce faulty data which are closer to the original non-faulty ones and hence likely to be more realistic. We experimentally investigated the effectiveness and cost of the introduced constraint negation strategies. The results provide insights for the application of these strategies in fault injection.

Distributed Acoustic Beamforming With Blockchain Protection

Speech is a natural user interface for the Internet of Things system. However, the presence of noise affects severely the performance of such system. With the deployment of smart devices with microphones, one can form a powerful acoustic sensor network to enhance the speech via beamforming techniques. On the other hand, reliability of data transmission also determines the beamforming performance, since faulty data will drift the beamformer steering location randomly. Currently, there is no protection scheme for acoustic data transmitted over the wireless network in order to keep steady beamforming performance. In this article, we design a compound distributed beamformer, where nodes are grouped and the system is embedded with blockchain technology to protect the data integrity during transmission. It attempts to provide more possible reliable connections between groups. Simulated experiments show that the distributed beamformer with blockchain protection is able to maintain steady beamforming performance.

A semi-supervised Support Vector Data Description-based fault detection method for rolling element bearings based on cyclic spectral analysis

The early and accurate detection of rolling element bearing faults, which is closely linked to the timely maintenance and repair before a sudden breakdown, is still one of the key challenges in the area of condition monitoring. Nowadays in the frames of research advanced signal processing techniques are combined with high level machine learning approaches, focusing towards automatic fault diagnosis and decision making. A plethora of Health Indicators (HIs) have been proposed to feed in machine learning models in orderto track the system degradation. Cyclic Spectral Analysis (CSA), including Cyclic Spectral Correlation (CSC) and Cyclic Spectral Coherence (CSCoh), has been proved as a powerful tool for rotating machinery fault detection. Due to the periodic mechanism of bearing fault impacts, the HIs extracted from the Cyclostationary (CS) domain can detect bearing defects even in premature stage. On the other hand, supervised machine learning approaches with labelled training and testing datasets cannot be yet realistically applied in industrial applications. In order to overcome this limitation, a novel semi-supervised Support Vector Data Description (SVDD) with negative samples (NSVDD) fault detection approach is proposed in this paper. The NSVDD model utilizes CS indicators to build the feature space, and fits a hyper-sphere to calculate the Euclidean distances in order to isolate the healthy and faulty data. An uniform object generation method is adopted to generate artificial outliers as negative samples for the NSVDD. A systematic fault detection decision strategy is proposed to estimate the bearing status simultaneously with the detection of fault initiation. Furthermore, a multi-level anomaly detection framework is built based on data at i) single sensor level, ii) machine level and iii) entire machine fleet level. Three run-to-failure bearing datasets including signals from twelve bearings are used to implement the proposed fault detection methodology. Results show that, the CS based indicators outperform time indicators and Fast Kurtogram (FK) based Squared Envelope Spectrum (SES) indicators. Moreover, the proposed NSVDD model show superior characteristics in anomaly detection compared to three classification methodologies, i.e. the Back-Propagation Neural Network, the Random Forest and the K-Nearest Neighbour.

Urban Traffic Data Imputation With Detrending and Tensor Decomposition

Due to various uncontrollable factors (such as random faulty acquisition equipment and data distortion), urban traffic flow data inevitably suffers from some form of data loss. Finding an effective filling method to estimate the missing data is of great help to the study of transportation networks. Traffic flow during a day are likely to have its regular peak period and off-peak period. For most regions of the urban road network, normally there is a certain trend in the traffic flow data. In this paper, we propose a data imputation method that employs a tensor decomposition approach, which fully considers the characteristics of the traffic flow in both time and space. The proposed method is based on high order singular value decomposition with soft thresholding core. In this method, traffic data are divided into its main trend part and the residual part, which is called detrending. And tensor decomposition is performed on these two parts separately. For each part, dynamic rank method is used to adjust the rank of tensor decomposition. With the actual 214 anonymous road segments with 10 minutes interval data in Guangdong, China, the highway data with 15 minutes interval in Madrid, Spain, and the traffic flow data from PeMS with 5 minutes interval in California, USA. The results of the different models are discussed in the case of continuous data missing and random data missing by different time intervals. In addition, by comparing with other data imputation methods, our method can fill the missing data with better performance.

Online Fault Detection in Autonomous Ferries: Using Fault-type Independent Spectral Anomaly Detection

Enthusiasm for ship autonomy is flourishing in the maritime industry. In this context, data-driven prognostics and health management (PHM) systems have emerged as the optimal way to improve operational reliability and system safety. However, further research is needed to enhance the essential actions relating to such a system. Fault detection is the first and most crucial action of any data-driven PHM system. In this article, we propose a fault-type independent spectral anomaly detection algorithm for marine diesel engine degradation in autonomous ferries. The benefits of the algorithm are verified on three fault types where the nature of degradation differs. Both normal operation data and faulty degradation data have been collected from a marine diesel engine using two different engine load profiles. These profiles aim to replicate real autonomous ferry crossing operations, environmental conditions that the ferry may encounter. First, the data are subjected to a feature selection process to remove irrelevant and redundant features. Then, a multiregime normalization method is performed on the data to merge the engine loads into one context. Finally, a variational autoencoder is trained to estimate velocity and acceleration calculations of the anomaly score. Generic and dynamic threshold limits are simultaneously established to detect the fault time step online. The algorithm achieved an accuracy of 97.66% in the final test when the acceleration was used as the fault detector. The results suggest that the algorithm is independent of fault types with different nature of degradation related to the marine diesel engine.

Measurement System Analysis and System Thinking in Six Sigma

This article investigated measurement system analysis and system thinking in Six Sigma, as well as the factors that influence these actions. If the measurement system being used to accumulate data from the process delivers dependable and accurate results, the measurement system analysis regulates it. Process improvement initiatives can be derailed by faulty measurement systems. Also, managers who have read faulty data can be misled into making wrong decisions. To collect trustworthy data, a reliable measurement system is established with this process. A method to assess an organization as a system and interpret its practices as a whole with Six Sigma is system thinking. Also, fixing a system as a whole helps to identify the real causes of issues and to know where to address them. This article addressed the contribution of these two methods to an overall success of an organization operating Six Sigma. The most current variables, concepts, and models were studied within operations and project management. By using a design-science-investigate strategy, this study approved of a valuable growth reveal for reasonable and hypothetical application. This study allowed us to generate a fitting assessment model that will fill the research void. Also, this study contributed to the engineering field with improved project success rates and team communication.

Fault Detection of Solar PV System Using SVM and Thermal Image Processing

Installation of photovoltaic plants across the globe increases, in the recent years, due to the energy demand across the world. Solar energy is free of cost, inexhaustible and a non-polluted source to the environment. The efficiency of any power plant depends on its fault free operation. Due to the occurrence of fault in PV system, the reliability and power output is reduced. In PV plants, the internal and external faults normally result in an increase in temperature which is easily sensed by different methods. In this paper, an algorithm based on thermal image processing, is proposed to extract the features of the PV cells in operation. These extracted features are then compared with the features of the healthy PV module using Support Vector Machine. SVM is a classifier tool which classifies whether the PV modules are defective or non-defective. An experimental set up is created and the performance of the algorithm is verified by testing it with faulty data sets which are obtained by creating different types of faults intentionally. The algorithm successfully identifies the defective PV module and its performance is validated experimentally.

Machine Learning Approach for Postprandial Blood Glucose Prediction in Gestational Diabetes Mellitus

Postprandial blood glucose prediction is a crucial part of diabetes management. Recently, this topic has been of great interest, resulting in many research projects and published papers. Although different input parameters that might be beneficial for blood glucose prediction models were comprehensively discussed, specific data preprocessing, feature engineering and model tuning steps were not explained in detail in many of these papers. In this work, we developed and comprehensively described a data-driven blood glucose model based on a decision tree gradient boosting algorithm to predict different characteristics of postprandial glycemic responses; the model utilized meal-related data derived from a mobile app diary (including information on the glycemic index), food context (information on previous meals), characteristics of the individual patients and patient behavioral questionnaires. A set of rules was defined and implemented to detect incorrect meal records and to filter faulty data, and analyses were conducted on the overall food diary data and in particular, the data on the current meal for which the postprandial blood glucose response was calculated. Different gradient boosting models were trained and evaluated with parameters selected via random search cross-validation. The best models for the prediction of the incremental area under the blood glucose curve two hours after food intake had the following characteristics: R = 0.631, MAE = 0.373 mmol/L*h for the model not using data on current blood glucose; R = 0.644, MAE = 0.371 mmol/L*h for the model using data on the current blood glucose levels; and R = 0.704, MAE = 0.341 mmol/L*h for the model utilizing data on the continuous blood glucose trends before the meal. The impact of features was evaluated using Shapley values. The meal glycemic load, amount of carbohydrates in the meal, type of meal (e.g., breakfast), amount of starch and amount of food consumed 6 hours before the current meal were the most important contributors in the models.

A Framework for Robust Assimilation of Potentially Malign Third-Party Data, and Its Statistical Meaning

This paper presents a model-based method for fusing data from multiple sensors with a hypothesis-test-based component for rejecting potentially faulty or otherwise malign data. Our framework is based on an extension of the classic particle filter algorithm for real-time state estimation of uncertain systems with nonlinear dynamics with partial and noisy observations. This extension, based on classical statistical theories, utilizes statistical tests against the system's observation model. We discuss the application of the two major statistical testing frameworks, Fisherian significance testing and Neyman-Pearsonian hypothesis testing, to the Monte Carlo and sensor fusion settings. The Monte Carlo Neyman-Pearson test we develop is useful when one has a reliable model of faulty data, while the Fisher one is applicable when one may not have a model of faults, which may occur when dealing with third-party data, like GNSS data of transportation system users. These statistical tests can be combined with a particle filter to obtain a Monte Carlo state estimation scheme that is robust to faulty or outlier data. We present a synthetic freeway traffic state estimation problem where the filters are able to reject simulated faulty GNSS measurements. The fault-model-free Fisher filter, while underperforming the Neyman-Pearson one when the latter has an accurate fault model, outperforms it when the assumed fault model is incorrect.

Detecting Gas Turbine Combustor Anomalies Using Semi-supervised Anomaly Detection with Deep Representation Learning

Deep learning (DL), regarded as a breakthrough machine learning technique, has proven to be effective for a variety of real-world applications. However, DL has not been actively applied to condition monitoring of industrial assets, such as gas turbine combustors. We propose a deep semi-supervised anomaly detection (deepSSAD) that has two key components: (1) using DL to learn representations or features from multivariate, time-series sensor measurements; and (2) using one-class classification to model normality in the learned feature space, thus performing anomaly detection. Both steps use normal data only; thus our anomaly detection falls into the semi-supervised anomaly detection category, which is advantageous for industrial asset condition monitoring where abnormal or faulty data is rare. Using the data collected from a real-world gas turbine combustion system, we demonstrate that our proposed approach achieved a good detection performance (AUC) of 0.9706 ± 0.0029. Furthermore, we compare the detection performance of the proposed approach against that of other different designs, including different features (i.e., the deep learned, handcrafted and PCA features) and different detection models (i.e., one-class ELM, one-class SVM, isolation forest, and Gaussian mixture model). The proposed approach significantly outperforms others. The proposed combustor anomaly detection approach is effective in detecting combustor anomalies or faults.

Unsupervised learning for fault detection and diagnosis of air handling units

Supervised learning techniques have witnessed significant successes in fault detection and diagnosis (FDD) for heating ventilation and air-conditioning (HVAC) systems. Despite the good performance, these techniques heavily rely on balanced datasets that contain a large amount of both faulty and normal data points. In real-world scenarios, however, it is often very challenging to collect a sufficient amount of faulty training samples that are necessary for building a balanced training dataset. In this paper, we introduce a framework that utilizes the generative adversarial network (GAN) to address the imbalanced data problem in FDD for air handling units (AHUs). To this end, we first show the necessary procedures of applying GAN to increase the number of faulty training samples in the training pool and re-balance the training dataset. The proposed framework then uses supervised classifiers to train the re-balanced datasets. Finally, we present a comparative study that illustrates the advantages of the proposed method for FDD of AHU with various evaluation metrics. Our work demonstrates the promising prospects of performing robust FDD of AHU with a limited number of faulty training samples.

Increasing sensor reliability through confidence attribution

The reliability of wireless sensor networks (WSN) is getting increasing importance as this kind of networks are becoming the communication base for many cyber-physical systems (CPS). Such systems rely on sensor data correctness to make decisions; therefore, faulty data can lead such systems to take wrong actions. Errors can be originated by sensor’s hardware failures or software bugs and also from the intentional interference of intruders. The gateways that connect such WSN to the Internet are natural intruders’ targets as they usually run conventional operating systems and communication protocols. This work proposes a confidence attribution scheme, based on lightweight predictors running on the sensors. The solution also proposes a parameterizable formula, in order to stamp every value sent by a sensor with a confidence level, calculated upon the values of a subset of correlated sensors. This work also presents an algorithm that can identify a defective sensor into its subset. The use of predictors and confidence attribution are proposed as the basis of a mechanism that increases the WSN resilience against sensor failure or bad data injection by intruders. Several simulations were performed to evaluate the detection efficiency against different types of sensor errors. This work also analyses mechanisms to deal with concept drifts in the WSN lifetime.

Faulty-data detection and data quality measure in cyber–physical systems through Weibull distribution

The preeminent challenge when data is compiled from cyber–physical system is the detection and filtering of faulty data. When the storage, processing, and communication of a reasonable amount of data migrate to those mobile and embedded devices, it becomes strenuous to apply conventional powerful algorithms for identifying, eliminating, and tolerating the faulty data because of the resource constraints of these devices. Further, the real-time requirements of CPS applications require CPSs to respond quickly and the aggregation method that is long-established in CPSs makes it challenging to ascertain the value of the original data. Thus, a systematic approach is essential to accomplish the goal of faulty data detection and filtering. In this paper, we present several approaches that we can deploy to detect and filter faulty data efficiently and cost-effectively, to improve the quality of the collected data from a system’s perspective to match the specific requirements of CPSs.