An Adaptive Machine Learning Decision System for Flexible Predictive Maintenance

Process monitoring and Predictive Maintenance (PdM) are gaining increasing attention in most manufacturing environments as a means of reducing maintenance related costs and downtime. This is especially true in industries that are data intensive such as semiconductor manufacturing. In this paper an adaptive PdM based flexible maintenance scheduling decision support system, which pays particular attention to associated opportunity and risk costs, is presented. The proposed system, which employs Machine Learning and regularized regression methods, exploits new information as it becomes available from newly processed components to refine remaining useful life estimates and associated costs and risks. The system has been validated on a real industrial dataset related to an Ion Beam Etching process for semiconductor manufacturing.

Stiffness Analysis and Experiment of a Novel 5-DoF Parallel Kinematic Machine Considering Gravitational Effects

In order to carry out high-precision machining of aerospace structural components with large size, thin wall and complex surface, this paper proposes a novel parallel kinematic machine (PKM) and formulates its semi-analytical theoretical stiffness model considering gravitational effects that is verified by stiffness experiments. From the viewpoint of topology structure, the novel PKM consists of two substructures in terms of the redundant and overconstrained parallel mechanisms that are connected by two interlinked revolute joints. The theoretical stiffness model of the novel PKM is established based upon the virtual work principle and deformation superposition principle after mapping the stiffness models of substructures from joint space to operated space by Jacobian matrices and considering the deformation contributions of interlinked revolute joints to two substructures. Meanwhile, the component gravities are treated as external payloads exerting on the end reference point of the novel PKM resorting to static equivalence principle. This approach is proved by comparing the theoretical stiffness values with experimental stiffness values in the same configurations, which also indicates equivalent gravity can be employed to describe the actual distributed gravities in an acceptable accuracy manner. Finally, on the basis of the verified theoretical stiffness model, the stiffness distributions of the novel PKM are illustrated and the contributions of component gravities to the stiffness of the novel PKM are discussed.

Discrete Conditional Phase-type Model Utilising a Multiclass Support Vector Machine for the Prediction of Retinopathy of Prematurity

Retinopathy of prematurity (ROP) is a rare disease in which retinal blood vessels of premature infants fail to develop normally, and is one of the major causes of childhood blindness throughout the world. The Discrete Conditional Phase-type (DC-Ph) model consists of two components, the conditional component measuring the inter-relationships between covariates and the survival component which models the survival distribution using a Coxian phase-type distribution. This paper expands the DC-Ph models by introducing a support vector machine (SVM), in the role of the conditional component. The SVM is capable of classifying multiple outcomes and is used to identify the infant's risk of developing ROP. Class imbalance makes predicting rare events difficult. A new class decomposition technique, which deals with the problem of multiclass imbalance, is introduced. Based on the SVM classification, the length of stay in the neonatal ward is modelled using a 5, 8 or 9 phase Coxian distribution.

Using Graph Components Derived from an Associative Concept Dictionary to Predict fMRI Neural Activation Patterns that Represent the Meaning of Nouns

In this study, we introduce an original distance definition for graphs, called the Markov-inverse-F measure (MiF). This measure enables the integration of classical graph theory indices with new knowledge pertaining to structural feature extraction from semantic networks. MiF improves the conventional Jaccard and/or Simpson indices, and reconciles both the geodesic information (random walk) and co-occurrence adjustment (degree balance and distribution). We measure the effectiveness of graph-based coefficients through the application of linguistic graph information for a neural activity recorded during conceptual processing in the human brain. Specifically, the MiF distance is computed between each of the nouns used in a previous neural experiment and each of the in-between words in a subgraph derived from the Edinburgh Word Association Thesaurus of English. From the MiF-based information matrix, a machine learning model can accurately obtain a scalar parameter that specifies the degree to which each voxel in (the MRI image of) the brain is activated by each word or each principal component of the intermediate semantic features. Furthermore, correlating the voxel information with the MiF-based principal components, a new computational neurolinguistics model with a network connectivity paradigm is created. This allows two dimensions of context space to be incorporated with both semantic and neural distributional representations.

Modelling the fresh properties of self-compacting concrete using support vector machine approach

The main objective of the study presented in this paper was to investigate the feasibility using support vector machines (SVM) for the prediction of the fresh properties of self-compacting concrete. The radial basis function (RBF) and polynomial kernels were used to predict these properties as a function of the content of mix components. The fresh properties were assessed with the slump flow, T50, T60, V-funnel time, Orimet time, and blocking ratio (L-box). The retention of these tests was also measured at 30 and 60 min after adding the first water. The water dosage varied from 188 to 208 L/m3, the dosage of superplasticiser (SP) from 3.8 to 5.8 kg/m3, and the volume of coarse aggregates from 220 to 360 L/m3. In total, twenty mixes were used to measure the fresh state properties with different mixture compositions. RBF kernel was more accurate compared to polynomial kernel based support vector machines with a root mean square error (RMSE) of 26.9 (correlation coefficient of R2 = 0.974) for slump flow prediction, a RMSE of 0.55 (R2 = 0.910) for T50 (s) prediction, a RMSE of 1.71 (R2 = 0.812) for T60 (s) prediction, a RMSE of 0.1517 (R2 = 0.990) for V-funnel time prediction, a RMSE of 3.99 (R2 = 0.976) for Orimet time prediction, and a RMSE of 0.042 (R2 = 0.988) for L-box ratio prediction, respectively. A sensitivity analysis was performed to evaluate the effects of the dosage of cement and limestone powder, the water content, the volumes of coarse aggregate and sand, the dosage of SP and the testing time on the predicted test responses. The analysis indicates that the proposed SVM RBF model can gain a high precision, which provides an alternative method for predicting the fresh properties of SCC.

Fractionation, chemical and toxicological characterization of tobacco smoke components

La fumée du tabac est un aérosol extrêmement complexe constitué de milliers de composés répartis entre la phase particulaire et la phase vapeur. Il a été démontré que les effets toxicologiques de cette fumée sont associés aux composés appartenant aux deux phases. Plusieurs composés biologiquement actifs ont été identifiés dans la fumée du tabac; cependant, il n’y a pas d’études démontrant la relation entre les réponses biologiques obtenues via les tests in vitro ou in vivo et les composés présents dans la fumée entière du tabac. Le but de la présente recherche est de développer des méthodes fiables et robustes de fractionnement de la fumée à l’aide de techniques de séparation analytique et de techniques de détection combinés à des essais in vitro toxicologiques. Une étude antérieure réalisée par nos collaborateurs a démontré que, suite à l’étude des produits de combustion de douze principaux composés du tabac, l’acide chlorogénique s’est avéré être le composé le plus cytotoxique selon les test in vitro du micronoyau. Ainsi, dans cette étude, une méthode par chromatographie préparative en phase liquide a été développée dans le but de fractionner les produits de combustion de l’acide chlorogénique. Les fractions des produits de combustion de l’acide chlorogénique ont ensuite été testées et les composés responsables de la toxicité de l’acide chlorogénique ont été identifiés. Le composé de la sous-fraction responsable en majeure partie de la cytoxicité a été identifié comme étant le catéchol, lequel fut confirmé par chromatographie en phase liquide/ spectrométrie de masse à temps de vol. Des études récentes ont démontré les effets toxicologiques de la fumée entière du tabac et l’implication spécifique de la phase vapeur. C’est pourquoi notre travail a ensuite été focalisé principalement à l’analyse de la fumée entière. La machine à fumer Borgwaldt RM20S® utilisée avec les chambres d’exposition cellulaire de British American Tobacco permettent l’étude in vitro de l’exposition de cellules à différentes concentrations de fumée entière du tabac. Les essais biologiques in vitro ont un degré élevé de variabilité, ainsi, il faut prendre en compte toutes les autres sources de variabilité pour évaluer avec précision la finalité toxicologique de ces essais; toutefois, la fiabilité de la génération de la fumée de la machine n’a jamais été évaluée jusqu’à maintenant. Nous avons donc déterminé la fiabilité de la génération et de la dilution (RSD entre 0,7 et 12 %) de la fumée en quantifiant la présence de deux gaz de référence (le CH4 par détection à ionisation de flamme et le CO par absorption infrarouge) et d’un composé de la phase particulaire, le solanesol (par chromatographie en phase liquide à haute performance). Ensuite, la relation entre la dose et la dilution des composés de la phase vapeur retrouvée dans la chambre d’exposition cellulaire a été caractérisée en utilisant une nouvelle technique d’extraction dite par HSSE (Headspace Stir Bar Sorptive Extraction) couplée à la chromatographie en phase liquide/ spectrométrie de masse. La répétabilité de la méthode a donné une valeur de RSD se situant entre 10 et 13 % pour cinq des composés de référence identifiés dans la phase vapeur de la fumée de cigarette. La réponse offrant la surface maximale d’aire sous la courbe a été obtenue en utilisant les conditions expérimentales suivantes : intervalle de temps d’exposition/ désorption de 10 0.5 min, température de désorption de 200°C pour 2 min et température de concentration cryogénique (cryofocussing) de -75°C. La précision de la dilution de la fumée est linéaire et est fonction de l’abondance des analytes ainsi que de la concentration (RSD de 6,2 à 17,2 %) avec des quantités de 6 à 450 ng pour les composés de référence. Ces résultats démontrent que la machine à fumer Borgwaldt RM20S® est un outil fiable pour générer et acheminer de façon répétitive et linéaire la fumée de cigarette aux cultures cellulaires in vitro. Notre approche consiste en l’élaboration d’une méthodologie permettant de travailler avec un composé unique du tabac, pouvant être appliqué à des échantillons plus complexes par la suite ; ex : la phase vapeur de la fumée de cigarette. La méthodologie ainsi développée peut potentiellement servir de méthode de standardisation pour l’évaluation d’instruments ou de l’identification de produits dans l’industrie de tabac.

Object Detection in Images by Components

In this paper we present a component based person detection system that is capable of detecting frontal, rear and near side views of people, and partially occluded persons in cluttered scenes. The framework that is described here for people is easily applied to other objects as well. The motivation for developing a component based approach is two fold: first, to enhance the performance of person detection systems on frontal and rear views of people and second, to develop a framework that directly addresses the problem of detecting people who are partially occluded or whose body parts blend in with the background. The data classification is handled by several support vector machine classifiers arranged in two layers. This architecture is known as Adaptive Combination of Classifiers (ACC). The system performs very well and is capable of detecting people even when all components of a person are not found. The performance of the system is significantly better than a full body person detector designed along similar lines. This suggests that the improved performance is due to the components based approach and the ACC data classification structure.

Machine learning and small robot navigation

Developing successful navigation and mapping strategies is an essential part of autonomous robot research. However, hardware limitations often make for inaccurate systems. This project serves to investigate efficient alternatives to mapping an environment, by first creating a mobile robot, and then applying machine learning to the robot and controlling systems to increase the robustness of the robot system. My mapping system consists of a semi-autonomous robot drone in communication with a stationary Linux computer system. There are learning systems running on both the robot and the more powerful Linux system. The first stage of this project was devoted to designing and building an inexpensive robot. Utilizing my prior experience from independent studies in robotics, I designed a small mobile robot that was well suited for simple navigation and mapping research. When the major components of the robot base were designed, I began to implement my design. This involved physically constructing the base of the robot, as well as researching and acquiring components such as sensors. Implementing the more complex sensors became a time-consuming task, involving much research and assistance from a variety of sources. A concurrent stage of the project involved researching and experimenting with different types of machine learning systems. I finally settled on using neural networks as the machine learning system to incorporate into my project. Neural nets can be thought of as a structure of interconnected nodes, through which information filters. The type of neural net that I chose to use is a type that requires a known set of data that serves to train the net to produce the desired output. Neural nets are particularly well suited for use with robotic systems as they can handle cases that lie at the extreme edges of the training set, such as may be produced by "noisy" sensor data. Through experimenting with available neural net code, I became familiar with the code and its function, and modified it to be more generic and reusable for multiple applications of neural nets.

Design, construction, and testing of a research quality electric machine testbed

With energy demands and costs growing every day, the need for improving energy efficiency in electrical devices has become very important. Research into various methods of improving efficiency for all electrical components will be a key to meet future energy needs. This report documents the design, construction, and testing of a research quality electric machine dynamometer and test bed. This test cell system can be used for research in several areas including: electric drives systems, electric vehicle propulsion systems, power electronic converters, load/source element in an AC Microgrid, as well as many others. The test cell design criteria, and decisions, will be discussed in reference to user functionality and flexibility. The individual power components will be discussed in detail to how they relate to the project, highlighting any feature used in operation of the test cell. A project timeline will be discussed, clearly stating the work done by the different individuals involved in the project. In addition, the system will be parameterized and benchmark data will be used to provide the functional operation of the system. With energy demands and costs growing every day, the need for improving energy efficiency in electrical devices has become very important. Research into various methods of improving efficiency for all electrical components will be a key to meet future energy needs. This report documents the design, construction, and testing of a research quality electric machine dynamometer and test bed. This test cell system can be used for research in several areas including: electric drives systems, electric vehicle propulsion systems, power electronic converters, load/source element in an AC Microgrid, as well as many others. The test cell design criteria, and decisions, will be discussed in reference to user functionality and flexibility. The individual power components will be discussed in detail to how they relate to the project, highlighting any feature used in operation of the test cell. A project timeline will be discussed, clearly stating the work done by the different individuals involved in the project. In addition, the system will be parameterized and benchmark data will be used to provide the functional operation of the system.

Shallow Dialogue Processing Using Machine Learning Algorithms (or not)

This paper presents a shallow dialogue analysis model, aimed at human-human dialogues in the context of staff or business meetings. Four components of the model are defined, and several machine learning techniques are used to extract features from dialogue transcripts: maximum entropy classifiers for dialogue acts, latent semantic analysis for topic segmentation, or decision tree classifiers for discourse markers. A rule-based approach is proposed for solving cross-modal references to meeting documents. The methods are trained and evaluated thanks to a common data set and annotation format. The integration of the components into an automated shallow dialogue parser opens the way to multimodal meeting processing and retrieval applications.

Optimization and Machine Learning Frameworks for Complex Network Analysis

Thesis (Ph.D.)--University of Washington, 2016-06

The influence of pressure during solidification of high pressure die cast aluminium telecommunications components

The effects of process variables on the quality of high-pressure die cast components was determined with the aid of in-cavity pressure sensors. In particular, the effects of set intensification pressure, delay time, and casting velocity have been investigated. The in-cavity pressure sensor has been used to determine how conditions within the die-cavity are related to the process parameters regulated by the die casting machine, and in turn the effect of variations in these parameters on the integrity of the final part. Porosity was found to decrease with increasing intensification pressure and increase with increasing casting velocity. The delay time before the application of the intensification pressure was not observed to have a significant effect on porosity levels. (c) 2006 Elsevier B.V. All rights reserved.

An integrated scheduling problem of PCB components on sequential pick-and-place machines:mathematical models and heuristic solutions

This paper formulates several mathematical models for determining the optimal sequence of component placements and assignment of component types to feeders simultaneously or the integrated scheduling problem for a type of surface mount technology placement machines, called the sequential pick-andplace (PAP) machine. A PAP machine has multiple stationary feeders storing components, a stationary working table holding a printed circuit board (PCB), and a movable placement head to pick up components from feeders and place them to a board. The objective of integrated problem is to minimize the total distance traveled by the placement head. Two integer nonlinear programming models are formulated first. Then, each of them is equivalently converted into an integer linear type. The models for the integrated problem are verified by two commercial packages. In addition, a hybrid genetic algorithm previously developed by the authors is adopted to solve the models. The algorithm not only generates the optimal solutions quickly for small-sized problems, but also outperforms the genetic algorithms developed by other researchers in terms of total traveling distance.

A heuristic approach for component scheduling on a high-speed PCB assembly machine

The collect-and-place machine is one of the most widely used placement machines for assembling electronic components on the printed circuit boards (PCBs). Nevertheless, the number of researches concerning the optimisation of the machine performance is very few. This motivates us to study the component scheduling problem for this type of machine with the objective of minimising the total assembly time. The component scheduling problem is an integration of the component sequencing problem, that is, the sequencing of component placements; and the feeder arrangement problem, that is, the assignment of component types to feeders. To solve the component scheduling problem efficiently, a hybrid genetic algorithm is developed in this paper. A numerical example is used to compare the performance of the algorithm with different component grouping approaches and different population sizes.

Uncertainty evaluation of an axi-symmetric measurement machine

This paper describes a method of uncertainty evaluation for axi-symmetric measurement machines which is compliant with GUM and PUMA methodologies. Specialized measuring machines for the inspection of axisymmetric components enable the measurement of properties such as roundness (radial runout), axial runout and coning. These machines typically consist of a rotary table and a number of contact measurement probes located on slideways. Sources of uncertainty include the probe calibration process, probe repeatability, probe alignment, geometric errors in the rotary table, the dimensional stability of the structure holding the probes and form errors in the reference hemisphere which is used to calibrate the system. The generic method is described and an evaluation of an industrial machine is described as a worked example. Type A uncertainties were obtained from a repeatability study of the probe calibration process, a repeatability study of the actual measurement process, a system stability test and an elastic deformation test. Type B uncertainties were obtained from calibration certificates and estimates. Expanded uncertainties, at 95% confidence, were then calculated for the measurement of; radial runout (1.2 µm with a plunger probe or 1.7 µm with a lever probe); axial runout (1.2 µm with a plunger probe or 1.5 µm with a lever probe); and coning/swash (0.44 arc seconds with a plunger probe or 0.60 arc seconds with a lever probe).