Search results for: failure detection and prediction

Authors: Maria L. G. Lourenço, Keylla H. N. P. Pereira, Viviane Y. Hibaru, Fabiana F. Souza, João C. P. Ferreira, Simone B. Chiacchio, Luiz H. A. Machado

Abstract:

Orphaned is a risk factor for mortality in newborns since it is a condition with total or partial absence of maternal care that is essential for neonatal survival, including nursing (nutrition, the transference of passive immunity and hydration), warmth, urination, and defecation stimuli, and protection. The most common causes of mortality in orphans are related to lack of assistance, handling mistakes and infections. This study aims to describe the orphans rates in neonatal puppies, the main causes, and the mortality rates. The study included 735 neonates admitted to the Sao Paulo State University (UNESP) Veterinary Hospital, Botucatu, Sao Paulo, Brazil, between January 2018 and November 2019. The orphans rate was 43.4% (319/735) of all neonates included, and the main causes for orphaned were related to maternal agalactia/hypogalactia (23.5%, 75/319); numerous litter (15.7%, 50/319), toxic milk syndrome due to maternal mastitis (14.4%, 46/319), absence of suction/weak neonate (12.2%, 39/319), maternal disease (9.4%, 30/319), cleft palate/lip (6.3%, 20/319), maternal death (5.9%, 19/319), prematurity (5.3%, 17/319), rejection/failure in maternal instinct (3.8%, 12/319) and abandonment by the owner/separation of mother and neonate (3.5%, 11/319). The main consequences of orphaned observed in the admitted neonates were hypoglycemia, hypothermia, dehydration, aspiration pneumonia, wasting syndrome, failure in the transference of passive immunity, infections and sepsis, which happened due to failure of identifying the problem early, lack of adequate assistance, negligence and handling mistakes by the owner. The total neonatal mortality rate was 8% (59/735) and the neonatal mortality rate among orphans was 18.5% (59/319). The orphaned and mortality rates were considered high, but even higher rates may be observed in locations without adequate neonatal assistance and owner orientation. The survival of these patients is related to constant monitoring of the litter, early diagnosis and assistance, and the implementation of effective handling for orphans. Understanding the correct handling for neonates and instructing the owners regarding proper handling are essential to minimize the consequences of orphaned and the mortality rates.

Keywords: orphans, neonatal care, puppies, newborn dogs

Procedia PDF Downloads 261

Authors: Jairson Barbosa Rodrigues, Paulo Romero Martins Maciel, Germano Crispim Vasconcelos

Abstract:

This paper presents an investigation of the performance impacts regarding the variation of five factors (input data size, node number, cores, memory, and disks) when applying a distributed implementation of Naïve Bayes for text classification of a large Corpus on the Spark big data processing framework. Problem: The algorithm's performance depends on multiple factors, and knowing before-hand the effects of each factor becomes especially critical as hardware is priced by time slice in cloud environments. Objectives: To explain the functional relationship between factors and performance and to develop linear predictor models for time and cost. Methods: the solid statistical principles of Design of Experiments (DoE), particularly the randomized two-level fractional factorial design with replications. This research involved 48 real clusters with different hardware arrangements. The metrics were analyzed using linear models for screening, ranking, and measurement of each factor's impact. Results: Our findings include prediction models and show some non-intuitive results about the small influence of cores and the neutrality of memory and disks on total execution time, and the non-significant impact of data input scale on costs, although notably impacts the execution time.

Keywords: big data, design of experiments, distributed machine learning, natural language processing, spark

Procedia PDF Downloads 123

Authors: Alireza Dantism

Abstract:

Presently, describing the function of a protein sequence is one of the most common problems in biology. Usually, this problem can be facilitated by studying the three-dimensional structure of proteins. In the absence of a protein structure, comparative modeling often provides a useful three-dimensional model of the protein that is dependent on at least one known protein structure. Comparative modeling predicts the three-dimensional structure of a given protein sequence (target) mainly based on its alignment with one or more proteins of known structure (templates). Comparative modeling consists of four main steps 1. Similarity between the target sequence and at least one known template structure 2. Alignment of target sequence and template(s) 3. Build a model based on alignment with the selected template(s). 4. Prediction of model errors 5. Optimization of the built model There are many computer programs and web servers that automate the comparative modeling process. One of the most important advantages of these servers is that it makes comparative modeling available to both experts and non-experts, and they can easily do their own modeling without the need for programming knowledge, but some other experts prefer using programming knowledge and do their modeling manually because by doing this they can maximize the accuracy of their modeling. In this study, a web-based tool has been designed to predict the tertiary structure of proteins using PHP and Python programming languages. This tool is called EasyModel. EasyModel can receive, according to the user's inputs, the desired unknown sequence (which we know as the target) in this study, the protein sequence file (template), etc., which also has a percentage of similarity with the primary sequence, and its third structure Predict the unknown sequence and present the results in the form of graphs and constructed protein files.

Keywords: structural bioinformatics, protein tertiary structure prediction, modeling, comparative modeling, modeller

Procedia PDF Downloads 102

Authors: Omar Dib, Rita Yaacoub, Luc Eveleigh, Nathalie Locquet, Hussein Dib, Ali Bassal, Christophe B. Y. Cordella

Abstract:

The potential of front-face fluorescence coupled with chemometric techniques, namely parallel factor analysis (PARAFAC) and multiple linear regression (MLR) as a rapid analysis tool to characterize Lebanese virgin olive oils was investigated. Fluorescence fingerprints were acquired directly on 102 Lebanese virgin olive oil samples in the range of 280-540 nm in excitation and 280-700 nm in emission. A PARAFAC model with seven components was considered optimal with a residual of 99.64% and core consistency value of 78.65. The model revealed seven main fluorescence profiles in olive oil and was mainly associated with tocopherols, polyphenols, chlorophyllic compounds and oxidation/hydrolysis products. 23 MLR regression models based on PARAFAC scores were generated, the majority of which showed a good correlation coefficient (R > 0.7 for 12 predicted variables), thus satisfactory prediction performances. Acid values, peroxide values, and Delta K had the models with the highest predictions, with R values of 0.89, 0.84 and 0.81 respectively. Among fatty acids, linoleic and oleic acids were also highly predicted with R values of 0.8 and 0.76, respectively. Factors contributing to the model's construction were related to common fluorophores found in olive oil, mainly chlorophyll, polyphenols, and oxidation products. This study demonstrates the interest of front-face fluorescence as a promising tool for quality control of Lebanese virgin olive oils.

Keywords: front-face fluorescence, Lebanese virgin olive oils, multiple Linear regressions, PARAFAC analysis

Procedia PDF Downloads 456

Authors: Michael Moeti, Khuliso Sigama, Thapelo Samuel Matlala

Abstract:

One of the principal challenges that the world is confronted with is insecurity. The crime rate is increasing exponentially, and protecting our physical assets especially in the motorist industry, is becoming impossible when applying our own strength. The need to develop technological solutions that detect and report theft without any human interference is inevitable. This is critical, especially for vehicle owners, to ensure theft detection and speedy identification towards recovery efforts in cases where a vehicle is missing or attempted theft is taking place. The vehicle theft detection system uses Convolutional Neural Network (CNN) to recognize the driver's face captured using an installed mobile phone device. The location identification function uses a Global Positioning System (GPS) to determine the real-time location of the vehicle. Upon identification of the location, Global System for Mobile Communications (GSM) technology is used to report or notify the vehicle owner about the whereabouts of the vehicle. The installed mobile app was implemented by making use of python as it is undoubtedly the best choice in machine learning. It allows easy access to machine learning algorithms through its widely developed library ecosystem. The graphical user interface was developed by making use of JAVA as it is better suited for mobile development. Google's online database (Firebase) was used as a means of storage for the application. The system integration test was performed using a simple percentage analysis. Sixty (60) vehicle owners participated in this study as a sample, and questionnaires were used in order to establish the acceptability of the system developed. The result indicates the efficiency of the proposed system, and consequently, the paper proposes the use of the system can effectively monitor the vehicle at any given place, even if it is driven outside its normal jurisdiction. More so, the system can be used as a database to detect, locate and report missing vehicles to different security agencies.

Keywords: CNN, location identification, tracking, GPS, GSM

Procedia PDF Downloads 176

Authors: Amaia Martinez Martinez, Martin Turek, Carlos Ventura, Jay Drew

Abstract:

This study aims to assess the seismic performance of arch and dome structural systems made from easy to assemble precast blocks of recycled concrete. These systems have been developed by Lock Block Ltd. Company from Vancouver, Canada, as an extension of their currently used retaining wall system. The characterization of the seismic behavior of these structures is performed by a combination of experimental static and dynamic testing, and analytical modeling. For the experimental testing, several tilt tests, as well as a program of shake table testing were undertaken using small scale arch models. A suite of earthquakes with different characteristics from important past events are chosen and scaled properly for the dynamic testing. Shake table testing applying the ground motions in just one direction (in the weak direction of the arch) and in the three directions were conducted and compared. The models were tested with increasing intensity until collapse occurred; which determines the failure level for each earthquake. Since the failure intensity varied with type of earthquake, a sensitivity analysis of the different parameters was performed, being impulses the dominant factor. For all cases, the arches exhibited the typical four-hinge failure mechanism, which was also shown in the analytical model. Experimental testing was also performed reinforcing the arches using a steel band over the structures anchored at both ends of the arch. The models were tested with different pretension levels. The bands were instrumented with strain gauges to measure the force produced by the shaking. These forces were used to develop engineering guidelines for the design of the reinforcement needed for these systems. In addition, an analytical discrete element model was created using 3DEC software. The blocks were designed as rigid blocks, assigning all the properties to the joints including also the contribution of the interlocking shear key between blocks. The model is calibrated to the experimental static tests and validated with the obtained results from the dynamic tests. Then the model can be used to scale up the results to the full scale structure and expanding it to different configurations and boundary conditions.

Keywords: arch, discrete element model, seismic assessment, shake-table testing

Procedia PDF Downloads 208

Authors: Muhammad Ahnaf Zahin, Yaw Adu-Gyamfi

Abstract:

Bus transit is a crucial component of transportation networks, especially in urban areas. Any intelligent transportation system must have accurate real-time information on bus travel times since it minimizes waiting times for passengers at different stations along a route, improves service reliability, and significantly optimizes travel patterns. Bus agencies must enhance the quality of their information service to serve their passengers better and draw in more travelers since people waiting at bus stops are frequently anxious about when the bus will arrive at their starting point and when it will reach their destination. For solving this issue, different models have been developed for predicting bus travel times recently, but most of them are focused on smaller road networks due to their relatively subpar performance in high-density urban areas on a vast network. This paper develops a deep learning-based architecture using a single-step multi-station forecasting approach to predict average bus travel times for numerous routes, stops, and trips on a large-scale network using heterogeneous bus transit data collected from the GTFS database. Over one week, data was gathered from multiple bus routes in Saint Louis, Missouri. In this study, Gated Recurrent Unit (GRU) neural network was followed to predict the mean vehicle travel times for different hours of the day for multiple stations along multiple routes. Historical time steps and prediction horizon were set up to 5 and 1, respectively, which means that five hours of historical average travel time data were used to predict average travel time for the following hour. The spatial and temporal information and the historical average travel times were captured from the dataset for model input parameters. As adjacency matrices for the spatial input parameters, the station distances and sequence numbers were used, and the time of day (hour) was considered for the temporal inputs. Other inputs, including volatility information such as standard deviation and variance of journey durations, were also included in the model to make it more robust. The model's performance was evaluated based on a metric called mean absolute percentage error (MAPE). The observed prediction errors for various routes, trips, and stations remained consistent throughout the day. The results showed that the developed model could predict travel times more accurately during peak traffic hours, having a MAPE of around 14%, and performed less accurately during the latter part of the day. In the context of a complicated transportation network in high-density urban areas, the model showed its applicability for real-time travel time prediction of public transportation and ensured the high quality of the predictions generated by the model.

Keywords: gated recurrent unit, mean absolute percentage error, single-step forecasting, travel time prediction.

Procedia PDF Downloads 76

Authors: Mohammad Raihan Mukhlis, M. Abdur Rahman Bhuiyan

Abstract:

Bangladesh is a developing country in which a lot of multi-span simply/continuous supported flyovers are being constructed in its major cities. Being situated in a seismically active region, seismic safety evaluation of flyovers is essential for seismic risk reduction. Effects of site amplification on seismic safety evaluation of flyover piers are the main concern of this study. In this regard, failure mode, lateral strength and displacement ductility of piers of a typical multi-span simply supported flyover have been evaluated by Japan Road Association (JRA) recommended guidelines, with and without considering site amplification. Ultimate flexural strengths of piers have been computed using the pushover analysis results. Shear capacity of piers has been calculated using the guidelines of JRA. Lateral strengths have been determined depending on the failure modes of the piers. Displacement ductility of piers has been computed using yield and ultimate displacements of the piers obtained from the pushover analysis results. Selected earthquake time history is used in seismic safety evaluation of the flyover piers. Finally, the ductility design method is used to conduct the seismic safety evaluation of the piers with and without considering site amplification. From the numerical results, it has been revealed that the effects of site amplification on seismic safety evaluation of bridge structures should be carefully taken into account.

Keywords: displacement ductility, flyover pier, lateral strength, safety evaluation, site amplification

Procedia PDF Downloads 173

Authors: Ali K. M. Al-Nasrawi, Uday A. Al-Hamdany, Sarah M. Hamylton, Brian G. Jones, Yasir M. Alyazichi

Abstract:

Surface elevation dynamics have always responded to disturbance regimes. Creating Digital Elevation Models (DEMs) to detect surface dynamics has led to the development of several methods, devices and data clouds. DEMs can provide accurate and quick results with cost efficiency, in comparison to the inherited geomatics survey techniques. Nowadays, remote sensing datasets have become a primary source to create DEMs, including LiDAR point clouds with GIS analytic tools. However, these data need to be tested for error detection and correction. This paper evaluates various DEMs from different data sources over time for Apple Orchard Island, a coastal site in southeastern Australia, in order to detect surface dynamics. Subsequently, 30 chosen locations were examined in the field to test the error of the DEMs surface detection using high resolution global positioning systems (GPSs). Results show significant surface elevation changes on Apple Orchard Island. Accretion occurred on most of the island while surface elevation loss due to erosion is limited to the northern and southern parts. Concurrently, the projected differential correction and validation method aimed to identify errors in the dataset. The resultant DEMs demonstrated a small error ratio (≤ 3%) from the gathered datasets when compared with the fieldwork survey using RTK-GPS. As modern modelling approaches need to become more effective and accurate, applying several tools to create different DEMs on a multi-temporal scale would allow easy predictions in time-cost-frames with more comprehensive coverage and greater accuracy. With a DEM technique for the eco-geomorphic context, such insights about the ecosystem dynamic detection, at such a coastal intertidal system, would be valuable to assess the accuracy of the predicted eco-geomorphic risk for the conservation management sustainability. Demonstrating this framework to evaluate the historical and current anthropogenic and environmental stressors on coastal surface elevation dynamism could be profitably applied worldwide.

Keywords: DEMs, eco-geomorphic-dynamic processes, geospatial Information Science, remote sensing, surface elevation changes,

Procedia PDF Downloads 269

Authors: Michael A. Kraus, Navid Pourmoghaddam, Martin Botz, Jens Schneider, Geralt Siebert

Abstract:

Fragmentation analysis of tempered glass gives insight into the quality of the tempering process and defines a certain degree of safety as well. Different standard such as the European EN 12150-1 or the American ASTM C 1048/CPSC 16 CFR 1201 define a minimum number of fragments required for soda-lime safety glass on the basis of fragmentation test results for classification. This work presents an approach for the glass breakage pattern prediction using a Voronoi Tesselation over Random Fields. The random Voronoi tessellation is trained with and validated against data from several breakage patterns. The fragments in observation areas of 50 mm x 50 mm were used for training and validation. All glass specimen used in this study were commercially available soda-lime glasses at three different thicknesses levels of 4 mm, 8 mm and 12 mm. The results of this work form a Bayesian framework for the training and prediction of breakage patterns of tempered soda-lime glass using a Voronoi Random Field Tesselation. Uncertainties occurring in this process can be well quantified, and several statistical measures of the pattern can be preservation with this method. Within this work it was found, that different Random Fields as basis for the Voronoi Tesselation lead to differently well fitted statistical properties of the glass breakage patterns. As the methodology is derived and kept general, the framework could be also applied to other random tesselations and crack pattern modelling purposes.

Keywords: glass breakage predicition, Voronoi Random Field Tessellation, fragmentation analysis, Bayesian parameter identification

Procedia PDF Downloads 163

Authors: I. Filikhin, B. Vlahovic

Abstract:

We focus on a novel type of detection based on electron tunneling properties of double nanoscale structures in semiconductor materials. Semiconductor heterostructures as quantum wells (QWs), quantum dots (QDs), and quantum rings (QRs) may have energy level structure of several hundred of electron confinement states. The single electron spectra of the double quantum objects (DQW, DQD, and DQR) were studied in our previous works with relation to the electron localization and tunneling between the objects. The wave function of electron may be localized in one of the QDs or be delocalized when it is spread over the whole system. The localizing-delocalizing tunneling occurs when an electron transition between both states is possible. The tunneling properties of spectra differ strongly for “regular” and “chaotic” systems. We have shown that a small violation of the geometry drastically affects localization of electron. In particular, such violations lead to the elimination of the delocalized states of the system. The same symmetry violation effect happens if electrical or magnetic fields are applied. These phenomena could be used to propose a new type of detection based on the high sensitivity of charge transport between double nanostructures and small violations of the shapes. It may have significant technological implications.

Keywords: double quantum dots, single electron levels, tunneling, electron localizations

Procedia PDF Downloads 509

Authors: Goodness Onwuka, Khaled Abou-El-Hossein

Abstract:

Borosilicate-crown (BK7) glass has found broad application in the optic and automotive industries and the growing demands for nanometric surface finishes is becoming a necessity in such applications. Thus, it has become paramount to optimize the parameters influencing the surface roughness of this precision lens. The research was carried out on a 4-axes Nanoform 250 precision lathe machine with an ultra-high precision grinding spindle. The experiment varied the machining parameters of feed rate, wheel speed and depth of cut at three levels for different combinations using Box Behnken design of experiment and the resulting surface roughness values were measured using a Taylor Hobson Dimension XL optical profiler. Acoustic emission monitoring technique was applied at a high sampling rate to monitor the machining process while further signal processing and feature extraction methods were implemented to generate the input to a neural network algorithm. This paper highlights the training and development of a back propagation neural network prediction algorithm through careful selection of parameters and the result show a better classification accuracy when compared to a previously developed response surface model with very similar machining parameters. Hence artificial neural network algorithms provide better surface roughness prediction accuracy in the ultra-high precision grinding of BK7 glass.

Keywords: acoustic emission technique, artificial neural network, surface roughness, ultra-high precision grinding

Procedia PDF Downloads 305

Authors: Naduni Ranasinghe

Abstract:

E-learning environments are getting more popular than any other due to the impact of COVID19. Even though e-learning is one of the best solutions for the teaching-learning process in the academic process, it’s not without major challenges. Nowadays, machine learning approaches are utilized in the analysis of how behavioral factors lead to better adoption and how they related to better performance of the students in eLearning environments. During the pandemic, we realized the academic process in the eLearning approach had a major issue, especially for the performance of the students. Therefore, an approach that investigates student behaviors in eLearning environments using a data-intensive machine learning approach is appreciated. A hybrid approach was used to understand how each previously told variables are related to the other. A more quantitative approach was used referred to literature to understand the weights of each factor for adoption and in terms of performance. The data set was collected from previously done research to help the training and testing process in ML. Special attention was made to incorporating different dimensionality of the data to understand the dependency levels of each. Five independent variables out of twelve variables were chosen based on their impact on the dependent variable, and by considering the descriptive statistics, out of three models developed (Random Forest classifier, SVM, and Decision tree classifier), random forest Classifier (Accuracy – 0.8542) gave the highest value for accuracy. Overall, this work met its goals of improving student performance by identifying students who are at-risk and dropout, emphasizing the necessity of using both static and dynamic data.

Keywords: academic performance prediction, e learning, learning analytics, machine learning, predictive model

Procedia PDF Downloads 161

Authors: Akila Murali, Prithika Manivel

Abstract:

Disasters are defined as a serious disruption of the functioning of a community or a society, which involves widespread human, material, economic or environmental impacts. The number of people suffering food crisis as a result of natural disasters has tripled in the last thirty years. The economic losses due to natural disasters have shown an increase with a factor of eight over the past four decades, caused by the increased vulnerability of the global society, and also due to an increase in the number of weather-related disasters. Efficient disaster detection and alerting systems could reduce the loss of life and properties. In the event of a disaster, another important issue is a good search and rescue system with high levels of precision, timeliness and safety for both the victims and the rescuers. Wireless Sensor Networks technology has the capability of quick capturing, processing, and transmission of critical data in real-time with high resolution. This paper studies the capacity of sensors and a Wireless Sensor Network to collect, collate and analyze valuable and worthwhile data, in an ordered manner to help with disaster management.

Keywords: alerting systems, disaster detection, Ad Hoc network, WSN technology

Procedia PDF Downloads 408

Authors: ShuJiang Jiang

Abstract:

The interaction of turbulence with airfoil is an important noise source in many engineering fields, including helicopters, turbofan, and contra-rotating open rotor engines, where turbulence generated in the wake of upstream blades interacts with the leading edge of downstream blades and produces aerodynamic noise. One approach to study turbulence-airfoil interaction noise is to model the oncoming turbulence as harmonic gusts. A compact noise source produces a dipole-like sound directivity pattern. However, when the acoustic wavelength is much smaller than the airfoil chord length, the airfoil needs to be treated as a non-compact source, and the gust-airfoil interaction becomes more complicated and results in multiple lobes generated in the radiated sound directivity. Capturing the short acoustic wavelength is a challenge for numerical simulations. In this work, simulations are performed for gust-airfoil interaction at different Mach numbers, using a high-fidelity direct Computational AeroAcoustic (CAA) approach based on a spectral/hp element method, verified by a CAA benchmark case. It is found that the squared sound pressure varies approximately as the 5th power of Mach number, which changes slightly with the observer location. This scaling law can give a better sound prediction than the flat-plate theory for thicker airfoils. Besides, another prediction method, based on the flat-plate theory and CAA simulation, has been proposed to give better predictions than the scaling law for thicker airfoils.

Keywords: aeroacoustics, gust-airfoil interaction, CFD, CAA

Procedia PDF Downloads 82

Authors: Kristel Dame Bañez Sumagaysay, Marie Victoria Cruz-javier

Abstract:

The current subject is a case of a 21 year-old woman at 29 1/7 weeks of gestation with Pre-B cell Acute Lymphoblastic Leukemia who was admitted to the coronary care unit (CCU) of the St. Luke’s Medical Center-Global City for Severe Acute Respiratory Distress Syndrome (ARDS) secondary to hospital-acquired pneumonia secondary to pneumocystis jiroveci; central line-associated bloodstream infection (E. aerogenes). She presented with chronic hypoxemia caused by Pulmonary edema, probably secondary to heart failure secondary to cardiomyopathy chemotherapy-induced. Due to worsening feto-maternal status, extracorporeal membrane oxygenation (ECMO) for respiratory support was instituted, and an elective cesarean section was done due to multiple maternal factors and deteriorating health status under total intravenous anesthesia assisted by veno-venous extracorporeal membrane oxygenation. She delivered a live preterm newborn male, APGAR Score: 1, 0, 0, birth weight 985 grams, birth length: 40.5cm, small for gestational age.

Keywords: extracorporeal membrane oxygenation, pre-b cell acute lymphoblastic leukemia, severe acute respiratory distress syndrome, ethical dilemmas

Procedia PDF Downloads 80

Authors: Julia Ive, Lucia Specia

Abstract:

Understanding and detecting toxicity is an important problem to support safer human interactions online. Our work focuses on the important problem of contextual toxicity detection, where automated classifiers are tasked with determining whether a short textual segment (usually a sentence) is toxic within its conversational context. We use “toxicity” as an umbrella term to denote a number of variants commonly named in the literature, including hate, abuse, offence, among others. Detecting toxicity in context is a non-trivial problem and has been addressed by very few previous studies. These previous studies have analysed the influence of conversational context in human perception of toxicity in controlled experiments and concluded that humans rarely change their judgements in the presence of context. They have also evaluated contextual detection models based on state-of-the-art Deep Learning and Natural Language Processing (NLP) techniques. Counterintuitively, they reached the general conclusion that computational models tend to suffer performance degradation in the presence of context. We challenge these empirical observations by devising better contextual predictive models that also rely on NLP data augmentation techniques to create larger and better data. In our study, we start by further analysing the human perception of toxicity in conversational data (i.e., tweets), in the absence versus presence of context, in this case, previous tweets in the same conversational thread. We observed that the conclusions of previous work on human perception are mainly due to data issues: The contextual data available does not provide sufficient evidence that context is indeed important (even for humans). The data problem is common in current toxicity datasets: cases labelled as toxic are either obviously toxic (i.e., overt toxicity with swear, racist, etc. words), and thus context does is not needed for a decision, or are ambiguous, vague or unclear even in the presence of context; in addition, the data contains labeling inconsistencies. To address this problem, we propose to automatically generate contextual samples where toxicity is not obvious (i.e., covert cases) without context or where different contexts can lead to different toxicity judgements for the same tweet. We generate toxic and non-toxic utterances conditioned on the context or on target tweets using a range of techniques for controlled text generation(e.g., Generative Adversarial Networks and steering techniques). On the contextual detection models, we posit that their poor performance is due to limitations on both of the data they are trained on (same problems stated above) and the architectures they use, which are not able to leverage context in effective ways. To improve on that, we propose text classification architectures that take the hierarchy of conversational utterances into account. In experiments benchmarking ours against previous models on existing and automatically generated data, we show that both data and architectural choices are very important. Our model achieves substantial performance improvements as compared to the baselines that are non-contextual or contextual but agnostic of the conversation structure.

Keywords: contextual toxicity detection, data augmentation, hierarchical text classification models, natural language processing

Procedia PDF Downloads 174

Authors: Abrahão S. Fontes, Carlos A. V. Cardoso, Levi P. B. Oliveira

Abstract:

The induction motors are used worldwide in various industries. Several maintenance techniques are applied to increase the operating time and the lifespan of these motors. Among these, the predictive maintenance techniques such as Motor Current Signature Analysis (MCSA), Motor Square Current Signature Analysis (MSCSA), Park's Vector Approach (PVA) and Park's Vector Square Modulus (PVSM) are used to detect and diagnose faults in electric motors, characterized by patterns in the stator current frequency spectrum. In this article, these techniques are applied and compared on a real motor, which has the fault of eccentricity in the air-gap. It was used as a theoretical model of an electric induction motor without fault in order to assist comparison between the stator current frequency spectrum patterns with and without faults. Metrics were purposed and applied to evaluate the sensitivity of each technique fault detection. The results presented here show that the above techniques are suitable for the fault of eccentricity in the air gap, whose comparison between these showed the suitability of each one.

Keywords: eccentricity in the air-gap, fault diagnosis, induction motors, predictive maintenance

Procedia PDF Downloads 352

Authors: Jatziri Y. Moreno-Martínez, Arturo Galván, Israel Enrique Herrera Díaz, José Ramón Gasca Tirado

Abstract:

In recent years, for the elevated viaducts in Mexico City, a construction system based on precast/pre-stressed concrete elements has been used, in which the bridge girders are divided in two parts by imposing a hinged support in sections where the bending moments that are originated by the gravity loads in a continuous beam are minimal. Precast concrete girders with dapped ends are a representative sample of a behavior that has complex configurations of stresses that make them more vulnerable to cracking due to flexure–shear interaction. The design procedures for ends of the dapped girders are well established and are based primarily on experimental tests performed for different configurations of reinforcement. The critical failure modes that can govern the design have been identified, and for each of them, the methods for computing the reinforcing steel that is needed to achieve adequate safety against failure have been proposed. Nevertheless, the design recommendations do not include procedures for controlling diagonal cracking at the entrant corner under service loading. These cracks could cause water penetration and degradation because of the corrosion of the steel reinforcement. The lack of visual access to the area makes it difficult to detect this damage and take timely corrective actions. Three-dimensional non-linear numerical models based on Finite Element Method to study the cracking at the entrant corner of dapped-end beams were performed using the software package ANSYS v. 11.0. The cracking was numerically simulated by using the smeared crack approach. The concrete structure was modeled using three-dimensional solid elements SOLID65 capable of cracking in tension and crushing in compression. Drucker-Prager yield surface was used to include the plastic deformations. The longitudinal post-tension was modeled using LINK8 elements with multilinear isotropic hardening behavior using von Misses plasticity. The reinforcement was introduced with smeared approach. The numerical models were calibrated using experimental tests carried out in “Instituto de Ingeniería, Universidad Nacional Autónoma de México”. In these numerical models the characteristics of the specimens were considered: typical solution based on vertical stirrups (hangers) and on vertical and horizontal hoops with a post-tensioned steel which contributed to a 74% of the flexural resistance. The post-tension is given by four steel wires with a 5/8’’ (16 mm) diameter. Each wire was tensioned to 147 kN and induced an average compressive stress of 4.90 MPa on the concrete section of the dapped end. The loading protocol consisted on applying symmetrical loading to reach the service load (180 kN). Due to the good correlation between experimental and numerical models some additional numerical models were proposed by considering different percentages of post-tension in order to find out how much it influences in the appearance of the cracking in the reentrant corner of the dapped-end beams. It was concluded that the increasing of percentage of post-tension decreases the displacements and the cracking in the reentrant corner takes longer to appear. The authors acknowledge at “Universidad de Guanajuato, Campus Celaya-Salvatierra” and the financial support of PRODEP-SEP (UGTO-PTC-460) of the Mexican government. The first author acknowledges at “Instituto de Ingeniería, Universidad Nacional Autónoma de México”.

Keywords: concrete dapped-end beams, cracking control, finite element analysis, postension

Procedia PDF Downloads 231

Authors: Leydi Noemi Peraza Gómez, Jose Álvarez Nemegyei, Damaris Francis Estrella Castillo

Abstract:

In order to have an epidemiological tool to detect biomechanical stress (ERGO-Mex), which impose physical labor or recreational activities, a questionnaire is constructed in Spanish, validated and culturally adapted to the Mayan indigenous population of Yucatan. Through the seven steps proposed by Guillemin and Beaton the procedure was: initial translation, synthesis of the translations, feed back of the translation. After that review by a committee of experts, pre-test of the preliminary version, and presentation of the results to the committee of experts and members of the community. Finally the evaluation of its internal validity (Cronbach's α coefficient) and external (intraclass correlation coefficient). The results for the validation in Spanish indicated that 45% of the participants have biomechanical stress. The ERGO-Mex correlation was 0.69 (p <0.0001). Subjects with high biomechanical stress had a higher score than subjects with low biomechanical stress (17.4 ± 8.9 vs.9.8 ± 2.8, p = 0.003). The Cronbach's α coefficient was 0.92; and for validation in Cronbach's α maya it was 0.82 and CCI = 0.70 (95% CI: 0.58-0.79; p˂0.0001); ERGO-Mex is suitable for performing early detection of musculoskeletal diseases and helping to prevent disability.

Keywords: biomechanical stress, disability, musculoskeletal disorders, prevention

Procedia PDF Downloads 184

Authors: Kripesh Adhikari, Hamid Bouchachia, Hammadi Nait-Charif

Abstract:

Falls are one of the major causes of injury and death among elderly people aged 65 and above. A support system to identify such kind of abnormal activities have become extremely important with the increase in ageing population. Pose estimation is a challenging task and to add more to this, it is even more challenging when pose estimations are performed on challenging poses that may occur during fall. Location of the body provides a clue where the person is at the time of fall. This paper presents a vision-based tracking strategy where available joints are grouped into three different feature points depending upon the section they are located in the body. The three feature points derived from different joints combinations represents the upper region or head region, mid-region or torso and lower region or leg region. Tracking is always challenging when a motion is involved. Hence the idea is to locate the regions in the body in every frame and consider it as the tracking strategy. Grouping these joints can be beneficial to achieve a stable region for tracking. The location of the body parts provides a crucial information to distinguish normal activities from falls.

Keywords: fall detection, machine learning, deep learning, pose estimation, tracking

Procedia PDF Downloads 192

Authors: Atuhurra Jesse, N'guessan Yves-Roland Douha, Pabitra Lenka

Abstract:

The detection of diseases within plants has attracted a lot of attention from computer vision enthusiasts. Despite the progress made to detect diseases in many plants, there remains a research gap to train image classifiers to detect the cacao swollen shoot virus disease or CSSVD for short, pertinent to cacao plants. This gap has mainly been due to the unavailability of high quality labeled training data. Moreover, institutions have been hesitant to share their data related to CSSVD. To fill these gaps, image classifiers to detect CSSVD-infected cacao plants are presented in this study. The classifiers are based on VGG16, ResNet50 and Vision Transformer (ViT). The image classifiers are evaluated on a recently released and publicly accessible KaraAgroAI Cocoa dataset. The best performing image classifier, based on ResNet50, achieves 95.39\% precision, 93.75\% recall, 94.34\% F1-score and 94\% accuracy on only 20 epochs. There is a +9.75\% improvement in recall when compared to previous works. These results indicate that the image classifiers learn to identify cacao plants infected with CSSVD.

Keywords: CSSVD, image classification, ResNet50, vision transformer, KaraAgroAI cocoa dataset

Procedia PDF Downloads 108

Authors: T. Vikneshvaran, A. Aminudin, U. Alyaa Hashim, Waziralilah N. Fathiah, D. Shakirah Shukor

Abstract:

This paper presents the experimental study conducted on a structure of three-floor height building model. Most vibrations are undesirable and can cause damages to the buildings, machines and people all around us. The vibration wave from earthquakes, construction and winds have high potential to bring damage to the buildings. Excessive vibrations can result in structural and machinery failures. This failure is related to the human life and environment around it. The effect of vibration which causes failure and damage to the high rise buildings can be controlled in real life by implementing tuned mass damper (TMD) into the structure of the buildings. This research aims to study the effect and performance improvement achieved by applying TMD into the building structure. A structure model of three degrees of freedom (3DOF) is designed to demonstrate the performance of TMD to the designed model. The model designed is the physical representation of actual building structure in real life. It is constructed at a reduced scale and will be used for the experiment. Thus, the result obtained will be more accurate to compared with the real life effect. Based on the result from experimental study, by applying TMD to the structure model, the forces of vibration and the displacement mode of the building reduced. Thus, the reduced in vibration of the building helps to maintain the good condition of the building.

Keywords: degrees-of-freedom, displacement mode, natural frequency, tuned mass damper

Procedia PDF Downloads 342

Authors: Benedictus Asriparusa, Lathifah Al Hakimi, Aulia Husada

Abstract:

A large amount of energy is being wasted by the release of natural gas associated with the oil industry. This release interrupts the environment particularly atmosphere layer condition globally which contributes to global warming impact. This research presents an overview of the methods employed by researchers in PT. Chevron Pacific Indonesia in the Minas area to determine a new prediction method of measuring and reducing gas flaring and its emission. The method emphasizes advanced research which involved analytical studies, numerical studies, modeling, and computer simulations, amongst other techniques. A flaring system is the controlled burning of natural gas in the course of routine oil and gas production operations. This burning occurs at the end of a flare stack or boom. The combustion process releases emissions of greenhouse gases such as NO2, CO2, SO2, etc. This condition will affect the chemical composition of air and environment around the boundary layer mainly during transition season. Transition season in Indonesia is absolutely very difficult condition to predict its pattern caused by the difference of two air mass conditions. This paper research focused on transition season in 2013. A simulation to create the new pattern of the pollutants distribution is needed. This paper has outlines trends in gas flaring modeling and current developments to predict the dominant variables in the pollutants distribution. A Fluent model is used to simulate the distribution of pollutants gas coming out of the stack, whereas WRF model output is used to overcome the limitations of the analysis of meteorological data and atmospheric conditions in the study area. Based on the running model, the most influence factor was wind speed. The goal of the simulation is to predict the new pattern based on the time of fastest wind and slowest wind occurs for pollutants distribution. According to the simulation results, it can be seen that the fastest wind (last of March) moves pollutants in a horizontal direction and the slowest wind (middle of May) moves pollutants vertically. Besides, the design of flare stack in compliance according to EPA Oil and Gas Facility Stack Parameters likely shows pollutants concentration remains on the under threshold NAAQS (National Ambient Air Quality Standards).

Keywords: flare motion, new prediction, pollutants distribution, transition season, WRF model

Procedia PDF Downloads 559

Authors: Seonaid Hume, Gordon Dobie, Graeme West

Abstract:

Detecting and localizing radioactive sources is a necessity for safe and secure decommissioning of nuclear facilities. An important aspect for the management of the sort-and-segregation process is establishing the spatial distributions and quantities of the waste radionuclides, their type, corresponding activity, and ultimately classification for disposal. The data received from surveys directly informs decommissioning plans, on-site incident management strategies, the approach needed for a new cell, as well as protecting the workforce and the public. Manual classification of nuclear waste from a nuclear cell is time-consuming, expensive, and requires significant expertise to make the classification judgment call. Also, in-cell decommissioning is still in its relative infancy, and few techniques are well-developed. As with any repetitive and routine tasks, there is the opportunity to improve the task of classifying nuclear waste using autonomous systems. Hence, this paper proposes a new framework for the automatic classification of nuclear waste. This framework consists of five main stages; 3D spatial mapping and object detection, object classification, radiological mapping, source localisation based on gathered evidence and finally, waste classification. The first stage of the framework, 3D visual mapping, involves object detection from point cloud data. A review of related applications in other industries is provided, and recommendations for approaches for waste classification are made. Object detection focusses initially on cylindrical objects since pipework is significant in nuclear cells and indeed any industrial site. The approach can be extended to other commonly occurring primitives such as spheres and cubes. This is in preparation of stage two, characterizing the point cloud data and estimating the dimensions, material, degradation, and mass of the objects detected in order to feature match them to an inventory of possible items found in that nuclear cell. Many items in nuclear cells are one-offs, have limited or poor drawings available, or have been modified since installation, and have complex interiors, which often and inadvertently pose difficulties when accessing certain zones and identifying waste remotely. Hence, this may require expert input to feature match objects. The third stage, radiological mapping, is similar in order to facilitate the characterization of the nuclear cell in terms of radiation fields, including the type of radiation, activity, and location within the nuclear cell. The fourth stage of the framework takes the visual map for stage 1, the object characterization from stage 2, and radiation map from stage 3 and fuses them together, providing a more detailed scene of the nuclear cell by identifying the location of radioactive materials in three dimensions. The last stage involves combining the evidence from the fused data sets to reveal the classification of the waste in Bq/kg, thus enabling better decision making and monitoring for in-cell decommissioning. The presentation of the framework is supported by representative case study data drawn from an application in decommissioning from a UK nuclear facility. This framework utilises recent advancements of the detection and mapping capabilities of complex radiation fields in three dimensions to make the process of classifying nuclear waste faster, more reliable, cost-effective and safer.

Keywords: nuclear decommissioning, radiation detection, object detection, waste classification

Procedia PDF Downloads 205

Authors: Hongyu Chen, Li Jiang

Abstract:

Ubiquitous anomalies endanger the security of our system con- stantly. They may bring irreversible damages to the system and cause leakage of privacy. Thus, it is of vital importance to promptly detect these anomalies. Traditional supervised methods such as Decision Trees and Support Vector Machine (SVM) are used to classify normality and abnormality. However, in some case, the abnormal status are largely rarer than normal status, which leads to decision bias of these methods. Generative adversarial network (GAN) has been proposed to handle the case. With its strong generative ability, it only needs to learn the distribution of normal status, and identify the abnormal status through the gap between it and the learned distribution. Nevertheless, existing GAN-based models are not suitable to process data with discrete values, leading to immense degradation of detection performance. To cope with the discrete features, in this paper, we propose an efficient GAN-based model with specifically-designed loss function. Experiment results show that our model outperforms state-of-the-art models on discrete dataset and remarkably reduce the overhead.

Keywords: GAN, discrete feature, Wasserstein distance, multiple intermediate layers

Procedia PDF Downloads 131

Authors: Tatiana G. Smirnova, Stan G. Benjamin

Abstract:

Rapid Update Cycle (RUC) land surface model (LSM) was a land-surface component in several generations of operational weather prediction models at the National Center for Environment Prediction (NCEP) at the National Oceanic and Atmospheric Administration (NOAA). It was designed for short-range weather predictions with an emphasis on severe weather and originally was intentionally simple to avoid uncertainties from poorly known parameters. Nevertheless, the RUC LSM, when coupled with the hourly-assimilating atmospheric model, can produce a realistic evolution of time-varying soil moisture and temperature, as well as the evolution of snow cover on the ground surface. This result is possible only if the soil/vegetation/snow component of the coupled weather prediction model has sufficient skill to avoid long-term drift. RUC LSM was first implemented in the operational NCEP Rapid Update Cycle (RUC) weather model in 1998 and later in the Weather Research Forecasting Model (WRF)-based Rapid Refresh (RAP) and High-resolution Rapid Refresh (HRRR). Being available to the international WRF community, it was implemented in operational weather models in Austria, New Zealand, and Switzerland. Based on the feedback from the US weather service offices and the international WRF community and also based on our own validation, RUC LSM has matured over the years. Also, a sea-ice module was added to RUC LSM for surface predictions over the Arctic sea-ice. Other modifications include refinements to the snow model and a more accurate specification of albedo, roughness length, and other surface properties. At present, RUC LSM is being tested in the regional application of the Unified Forecast System (UFS). The next generation UFS-based regional Rapid Refresh FV3 Standalone (RRFS) model will replace operational RAP and HRRR at NCEP. Over time, RUC LSM participated in several international model intercomparison projects to verify its skill using observed atmospheric forcing. The ESM-SnowMIP was the last of these experiments focused on the verification of snow models for open and forested regions. The simulations were performed for ten sites located in different climatic zones of the world forced with observed atmospheric conditions. While most of the 26 participating models have more sophisticated snow parameterizations than in RUC, RUC LSM got a high ranking in simulations of both snow water equivalent and surface temperature. However, ESM-SnowMIP experiment also revealed some issues in the RUC snow model, which will be addressed in this paper. One of them is the treatment of grid cells partially covered with snow. RUC snow module computes energy and moisture budgets of snow-covered and snow-free areas separately by aggregating the solutions at the end of each time step. Such treatment elevates the importance of computing in the model snow cover fraction. Improvements to the original simplistic threshold-based approach have been implemented and tested both offline and in the coupled weather model. The detailed description of changes to the snow cover fraction and other modifications to RUC soil and snow parameterizations will be described in this paper.

Keywords: land-surface models, weather prediction, hydrology, boundary-layer processes

Procedia PDF Downloads 93

Authors: Sara ElElimy, Samir Moustafa

Abstract:

Mobile network operators start to face many challenges in the digital era, especially with high demands from customers. Since mobile network operators are considered a source of big data, traditional techniques are not effective with new era of big data, Internet of things (IoT) and 5G; as a result, handling effectively different big datasets becomes a vital task for operators with the continuous growth of data and moving from long term evolution (LTE) to 5G. So, there is an urgent need for effective Big data analytics to predict future demands, traffic, and network performance to full fill the requirements of the fifth generation of mobile network technology. In this paper, we introduce data science techniques using machine learning and deep learning algorithms: the autoregressive integrated moving average (ARIMA), Bayesian-based curve fitting, and recurrent neural network (RNN) are employed for a data-driven application to mobile network operators. The main framework included in models are identification parameters of each model, estimation, prediction, and final data-driven application of this prediction from business and network performance applications. These models are applied to Telecom Italia Big Data challenge call detail records (CDRs) datasets. The performance of these models is found out using a specific well-known evaluation criteria shows that ARIMA (machine learning-based model) is more accurate as a predictive model in such a dataset than the RNN (deep learning model).

Keywords: big data analytics, machine learning, CDRs, 5G

Procedia PDF Downloads 142

Authors: Soheila Sadeghi

Abstract:

Construction projects are complex and often subject to significant cost overruns due to the multifaceted nature of the activities involved. Accurate cost estimation is crucial for effective budget planning and resource allocation. Traditional methods for predicting overruns often rely on expert judgment or analysis of historical data, which can be time-consuming, subjective, and may fail to consider important factors. However, with the increasing availability of data from construction projects, machine learning techniques can be leveraged to improve the accuracy of overrun predictions. This study applied machine learning algorithms to enhance the prediction of cost overruns in a case study of a construction project. The methodology involved the development and evaluation of two machine learning models: Random Forest and Neural Networks. Random Forest can handle high-dimensional data, capture complex relationships, and provide feature importance estimates. Neural Networks, particularly Deep Neural Networks (DNNs), are capable of automatically learning and modeling complex, non-linear relationships between input features and the target variable. These models can adapt to new data, reduce human bias, and uncover hidden patterns in the dataset. The findings of this study demonstrate that both Random Forest and Neural Networks can significantly improve the accuracy of cost overrun predictions compared to traditional methods. The Random Forest model also identified key cost drivers and risk factors, such as changes in the scope of work and delays in material delivery, which can inform better project risk management. However, the study acknowledges several limitations. First, the findings are based on a single construction project, which may limit the generalizability of the results to other projects or contexts. Second, the dataset, although comprehensive, may not capture all relevant factors influencing cost overruns, such as external economic conditions or political factors. Third, the study focuses primarily on cost overruns, while schedule overruns are not explicitly addressed. Future research should explore the application of machine learning techniques to a broader range of projects, incorporate additional data sources, and investigate the prediction of both cost and schedule overruns simultaneously.

Keywords: cost prediction, machine learning, project management, random forest, neural networks

Procedia PDF Downloads 65

Authors: Rita Chattopadhyay, Vivek Anand Thoutam

Abstract:

Speech is the most acceptable means of communication where we can quickly exchange our feelings and thoughts. Quite often, people can communicate orally but cannot interact or work with computers or devices. It’s easy and quick to give speech commands than typing commands to computers. In the same way, it’s easy listening to audio played from a device than extract output from computers or devices. Especially with Robotics being an emerging market with applications in warehouses, the hospitality industry, consumer electronics, assistive technology, etc., speech-based human-machine interaction is emerging as a lucrative feature for robot manufacturers. Considering this factor, the objective of this paper is to design the “Audio-Visual Co-Data Processing Pipeline.” This pipeline is an integrated version of Automatic speech recognition, a Natural language model for text understanding, object detection, and text-to-speech modules. There are many Deep Learning models for each type of the modules mentioned above, but OpenVINO Model Zoo models are used because the OpenVINO toolkit covers both computer vision and non-computer vision workloads across Intel hardware and maximizes performance, and accelerates application development. A speech command is given as input that has information about target objects to be detected and start and end times to extract the required interval from the video. Speech is converted to text using the Automatic speech recognition QuartzNet model. The summary is extracted from text using a natural language model Generative Pre-Trained Transformer-3 (GPT-3). Based on the summary, essential frames from the video are extracted, and the You Only Look Once (YOLO) object detection model detects You Only Look Once (YOLO) objects on these extracted frames. Frame numbers that have target objects (specified objects in the speech command) are saved as text. Finally, this text (frame numbers) is converted to speech using text to speech model and will be played from the device. This project is developed for 80 You Only Look Once (YOLO) labels, and the user can extract frames based on only one or two target labels. This pipeline can be extended for more than two target labels easily by making appropriate changes in the object detection module. This project is developed for four different speech command formats by including sample examples in the prompt used by Generative Pre-Trained Transformer-3 (GPT-3) model. Based on user preference, one can come up with a new speech command format by including some examples of the respective format in the prompt used by the Generative Pre-Trained Transformer-3 (GPT-3) model. This pipeline can be used in many projects like human-machine interface, human-robot interaction, and surveillance through speech commands. All object detection projects can be upgraded using this pipeline so that one can give speech commands and output is played from the device.

Keywords: OpenVINO, automatic speech recognition, natural language processing, object detection, text to speech

Procedia PDF Downloads 82