In a Bayesian learning setting, the posterior distribution of a predictive model arises from a tr... more In a Bayesian learning setting, the posterior distribution of a predictive model arises from a trade-off between its prior distribution and the conditional likelihood of observed data. Such distribution functions usually rely on additional hyperparameters which need to be tuned in order to achieve optimum predictive performance; this operation can be efficiently performed in an Empirical Bayes fashion by maximizing the posterior marginal likelihood of the observed data. Since the score function of this optimization problem is in general characterized by the presence of local optima, it is necessary to resort to global optimization strategies, which require a large number of function evaluations. Given that the evaluation is usually computationally intensive and badly scaled with respect to the dataset size, the maximum number of observations that can be treated simultaneously is quite limited. In this paper, we consider the case of hyperparameter tuning in Gaussian process regression. A straightforward implementation of the posterior log-likelihood for this model requires O(N^3) operations for every iteration of the optimization procedure, where N is the number of examples in the input dataset. We derive a novel set of identities that allow, after an initial overhead of O(N^3), the evaluation of the score function, as well as the Jacobian and Hessian matrices, in O(N) operations. We prove how the proposed identities, that follow from the eigendecomposition of the kernel matrix, yield a reduction of several orders of magnitude in the computation time for the hyperparameter optimization problem. Notably, the proposed solution provides computational advantages even with respect to state of the art approximations that rely on sparse kernel matrices.
2013 IEEE International Conference on Automation Science and Engineering (CASE), 2013
ABSTRACT In semiconductor fabrication processes, effective management of maintenance operations i... more ABSTRACT In semiconductor fabrication processes, effective management of maintenance operations is fundamental to decrease costs associated with failures and downtime. Predictive Maintenance (PdM) approaches, based on statistical methods and historical data, are becoming popular for their predictive capabilities and low (potentially zero) added costs. We present here a PdM module based on Support Vector Machines for prediction of integral type faults, that is, the kind of failures that happen due to machine usage and stress of equipment parts. The proposed module may also be employed as a health factor indicator. The module has been applied to a frequent maintenance problem in semiconductor manufacturing industry, namely the breaking of the filament in the ion-source of ion-implantation tools. The PdM has been tested on a real production dataset.
ABSTRACT Motivation In many cases, since it is hard to track all the variables and subsystems rel... more ABSTRACT Motivation In many cases, since it is hard to track all the variables and subsystems relevant to the process, results and effects of any anomaly are detected only retrospectively. Hence, the need for refined controls. Description of the Approach In our project, we approach the FDC on flow gases measurement through a simple mdel by analyzing �available parameters and measurement systems (like OES or ICP) with advenced statistical techniques. Evaluation of Results With this approach, we introduce a system that detects in "real time", at wafer-to-wafer or lot-to-lot level, any fault of the mfc subsystem controlling precursor gases flow in the process. In this way we will be able to reduce wafer scrap, equipment down time and engineer effort. ��������� � ����������� � ����� ������� ������������������ �������� ���������� ���� ���� � ��������� ����� �� �� ���������� �� ������������������� �� �� � �� �������� ����� ���� ����� �� �� �� ������� ��� �� ��������� ! ���� ������ �� � ����� � ����������� �� � �������� �� � ����� � �� � �� �� � ���� � ���� ���! �������� ���� ����������"��!��������� � ����� � �������� ���� � ������� ���#��� �$%&����� '(� �� � ������ ��������������� � ��)� �� ���������� �� " ����� � * �� �� ���������� � ��������� ������� ��� ���� � �������+� ������ + � ��� �� ����� �� ����������������� � � ��������������� �, ��������� �� ����������!��� �������!�� ����� ����� ����� �������� ��� ��� � ���� � ���� ����� ��� � �� ������� � )���� ����� ����� ����� �!
In this study, a gas-chromatography mass spectrometry (GC-MS) metabolomics study was applied to e... more In this study, a gas-chromatography mass spectrometry (GC-MS) metabolomics study was applied to examine urine metabolite profiles of different classes of neonates under different nutrition regimens. The study population included 35 neonates, exclusively either breastfed or formula milk fed, in a seven-day timeframe. Urine samples were collected from intrauterine growth restriction (IUGR), large for gestational age (LGA), and appropriate gestational age (AGA) neonates. At birth, IUGR and LGA neonates showed similarities in their urine metabolite profiles that differed from AGA. When neonates started milk feeding, their metabolite excretion profile was strongly characterized by the different diet regimens. After three days of formula milk nutrition, urine had higher levels of glucose, galactose, glycine and myo-inositol, while up-regulated aconitic acid, aminomalonic acid and adipic acid were found in breast milk fed neonates. At seven days, neonates fed with formula milk shared higher levels of pseudouridine with IUGR and LGA at birth. Breastfed neonates shared up-regulated pyroglutamic acid, citric acid, and homoserine, with AGA at birth. The role of most important metabolites is herein discussed.
Adipose tissue is no longer considered as inert; the literature describes the role it plays in th... more Adipose tissue is no longer considered as inert; the literature describes the role it plays in the production of many substances, such as adiponectin, visfatin, ghrelin, S100B, apelin, TNF, IL-6 and leptin. These molecules have specific roles in humans and their potential as biomarkers useful for identifying alterations related to intrauterine growth retardation and large for gestational age neonates is emerging. Infants born in such conditions have undergone metabolic changes, such as fetal hypo- or hyperinsulinemia, which may lead to development of dysmetabolic syndrome and other chronic diseases in adulthood. In this review, these biomarkers are analyzed specifically and it is discussed how metabolomics may be an advantageous tool for detection, discrimination and prediction of metabolic alterations and diseases. Thus, a holistic approach, such as metabolomics, could help the prevention and early diagnosis of metabolic syndrome.
2010 IEEE International Conference on Automation Science and Engineering, 2010
AbstractPredictive Maintenance methods are aimed to obtain reliable estimates of the remaining l... more AbstractPredictive Maintenance methods are aimed to obtain reliable estimates of the remaining life cycle of an equipment from time series of suitable process parameters, named health factors, typically exhibiting a monotone evo-lution associated with the equipment wear. ...
2012 IEEE International Conference on Automation Science and Engineering (CASE), 2012
Abstract In semiconductor manufacturing, state of the art for wafer quality control relies on pro... more Abstract In semiconductor manufacturing, state of the art for wafer quality control relies on product monitoring and feedback control loops; the involved metrology operations are particularly cost-intensive and time-consuming. For this reason, it is a common practice to measure a small subset of a productive lot and devoted to represent the whole lot. Virtual Metrology (VM) methodologies are able to obtain reliable predictions of metrology results at process time; this goal is usually achieved by means of statistical models, linking process ...
ABSTRACT In this paper, a multiple classifier machine learning (ML) methodology for predictive ma... more ABSTRACT In this paper, a multiple classifier machine learning (ML) methodology for predictive maintenance (PdM) is presented. PdM is a prominent strategy for dealing with maintenance issues given the increasing need to minimize downtime and associated costs. One of the challenges with PdM is generating the so-called “health factors,” or quantitative indicators, of the status of a system associated with a given maintenance issue, and determining their relationship to operating costs and failure risk. The proposed PdM methodology allows dynamical decision rules to be adopted for maintenance management, and can be used with high-dimensional and censored data problems. This is achieved by training multiple classification modules with different prediction horizons to provide different performance tradeoffs in terms of frequency of unexpected breaks and unexploited lifetime, and then employing this information in an operating cost-based maintenance decision system to minimize expected costs. The effectiveness of the methodology is demonstrated using a simulated example and a benchmark semiconductor manufacturing maintenance problem.
ABSTRACT Smart management of maintenances has become fundamental in manufacturing environments in... more ABSTRACT Smart management of maintenances has become fundamental in manufacturing environments in order to decrease downtime and costs associated with failures. Predictive Maintenance (PdM) systems based on Machine Learning (ML) techniques have the possibility with low added costs of drastically decrease failures-related expenses; given the increase of availability of data and capabilities of ML tools, PdM systems are becoming really popular, especially in semiconductor manufacturing. A PdM module based on Classification methods is presented here for the prediction of integral type faults that are related to machine usage and stress of equipment parts. The module has been applied to an important class of semiconductor processes, ion-implantation, for the prediction of ion-source tungsten filament breaks. The PdM has been tested on a real production dataset.
2011 IEEE International Conference on Automation Science and Engineering, 2011
The present paper is motivated by the application of Predictive Maintenance (PM) techniques in th... more The present paper is motivated by the application of Predictive Maintenance (PM) techniques in the semiconduc- tor manufacturing environment: such techniques are able, using process data, to make reliable predictions of residual equipment lifetime. The employment of PM yields positive fallouts on the productive process in form of unscheduled downtime reduction, increased spare parts availability and improved overall pro- duction
In a Bayesian learning setting, the posterior distribution of a predictive model arises from a tr... more In a Bayesian learning setting, the posterior distribution of a predictive model arises from a trade-off between its prior distribution and the conditional likelihood of observed data. Such distribution functions usually rely on additional hyperparameters which need to be tuned in order to achieve optimum predictive performance; this operation can be efficiently performed in an Empirical Bayes fashion by maximizing the posterior marginal likelihood of the observed data. Since the score function of this optimization problem is in general characterized by the presence of local optima, it is necessary to resort to global optimization strategies, which require a large number of function evaluations. Given that the evaluation is usually computationally intensive and badly scaled with respect to the dataset size, the maximum number of observations that can be treated simultaneously is quite limited. In this paper, we consider the case of hyperparameter tuning in Gaussian process regression. A straightforward implementation of the posterior log-likelihood for this model requires O(N^3) operations for every iteration of the optimization procedure, where N is the number of examples in the input dataset. We derive a novel set of identities that allow, after an initial overhead of O(N^3), the evaluation of the score function, as well as the Jacobian and Hessian matrices, in O(N) operations. We prove how the proposed identities, that follow from the eigendecomposition of the kernel matrix, yield a reduction of several orders of magnitude in the computation time for the hyperparameter optimization problem. Notably, the proposed solution provides computational advantages even with respect to state of the art approximations that rely on sparse kernel matrices.
2013 IEEE International Conference on Automation Science and Engineering (CASE), 2013
ABSTRACT In semiconductor fabrication processes, effective management of maintenance operations i... more ABSTRACT In semiconductor fabrication processes, effective management of maintenance operations is fundamental to decrease costs associated with failures and downtime. Predictive Maintenance (PdM) approaches, based on statistical methods and historical data, are becoming popular for their predictive capabilities and low (potentially zero) added costs. We present here a PdM module based on Support Vector Machines for prediction of integral type faults, that is, the kind of failures that happen due to machine usage and stress of equipment parts. The proposed module may also be employed as a health factor indicator. The module has been applied to a frequent maintenance problem in semiconductor manufacturing industry, namely the breaking of the filament in the ion-source of ion-implantation tools. The PdM has been tested on a real production dataset.
ABSTRACT Motivation In many cases, since it is hard to track all the variables and subsystems rel... more ABSTRACT Motivation In many cases, since it is hard to track all the variables and subsystems relevant to the process, results and effects of any anomaly are detected only retrospectively. Hence, the need for refined controls. Description of the Approach In our project, we approach the FDC on flow gases measurement through a simple mdel by analyzing �available parameters and measurement systems (like OES or ICP) with advenced statistical techniques. Evaluation of Results With this approach, we introduce a system that detects in "real time", at wafer-to-wafer or lot-to-lot level, any fault of the mfc subsystem controlling precursor gases flow in the process. In this way we will be able to reduce wafer scrap, equipment down time and engineer effort. ��������� � ����������� � ����� ������� ������������������ �������� ���������� ���� ���� � ��������� ����� �� �� ���������� �� ������������������� �� �� � �� �������� ����� ���� ����� �� �� �� ������� ��� �� ��������� ! ���� ������ �� � ����� � ����������� �� � �������� �� � ����� � �� � �� �� � ���� � ���� ���! �������� ���� ����������"��!��������� � ����� � �������� ���� � ������� ���#��� �$%&����� '(� �� � ������ ��������������� � ��)� �� ���������� �� " ����� � * �� �� ���������� � ��������� ������� ��� ���� � �������+� ������ + � ��� �� ����� �� ����������������� � � ��������������� �, ��������� �� ����������!��� �������!�� ����� ����� ����� �������� ��� ��� � ���� � ���� ����� ��� � �� ������� � )���� ����� ����� ����� �!
In this study, a gas-chromatography mass spectrometry (GC-MS) metabolomics study was applied to e... more In this study, a gas-chromatography mass spectrometry (GC-MS) metabolomics study was applied to examine urine metabolite profiles of different classes of neonates under different nutrition regimens. The study population included 35 neonates, exclusively either breastfed or formula milk fed, in a seven-day timeframe. Urine samples were collected from intrauterine growth restriction (IUGR), large for gestational age (LGA), and appropriate gestational age (AGA) neonates. At birth, IUGR and LGA neonates showed similarities in their urine metabolite profiles that differed from AGA. When neonates started milk feeding, their metabolite excretion profile was strongly characterized by the different diet regimens. After three days of formula milk nutrition, urine had higher levels of glucose, galactose, glycine and myo-inositol, while up-regulated aconitic acid, aminomalonic acid and adipic acid were found in breast milk fed neonates. At seven days, neonates fed with formula milk shared higher levels of pseudouridine with IUGR and LGA at birth. Breastfed neonates shared up-regulated pyroglutamic acid, citric acid, and homoserine, with AGA at birth. The role of most important metabolites is herein discussed.
Adipose tissue is no longer considered as inert; the literature describes the role it plays in th... more Adipose tissue is no longer considered as inert; the literature describes the role it plays in the production of many substances, such as adiponectin, visfatin, ghrelin, S100B, apelin, TNF, IL-6 and leptin. These molecules have specific roles in humans and their potential as biomarkers useful for identifying alterations related to intrauterine growth retardation and large for gestational age neonates is emerging. Infants born in such conditions have undergone metabolic changes, such as fetal hypo- or hyperinsulinemia, which may lead to development of dysmetabolic syndrome and other chronic diseases in adulthood. In this review, these biomarkers are analyzed specifically and it is discussed how metabolomics may be an advantageous tool for detection, discrimination and prediction of metabolic alterations and diseases. Thus, a holistic approach, such as metabolomics, could help the prevention and early diagnosis of metabolic syndrome.
2010 IEEE International Conference on Automation Science and Engineering, 2010
AbstractPredictive Maintenance methods are aimed to obtain reliable estimates of the remaining l... more AbstractPredictive Maintenance methods are aimed to obtain reliable estimates of the remaining life cycle of an equipment from time series of suitable process parameters, named health factors, typically exhibiting a monotone evo-lution associated with the equipment wear. ...
2012 IEEE International Conference on Automation Science and Engineering (CASE), 2012
Abstract In semiconductor manufacturing, state of the art for wafer quality control relies on pro... more Abstract In semiconductor manufacturing, state of the art for wafer quality control relies on product monitoring and feedback control loops; the involved metrology operations are particularly cost-intensive and time-consuming. For this reason, it is a common practice to measure a small subset of a productive lot and devoted to represent the whole lot. Virtual Metrology (VM) methodologies are able to obtain reliable predictions of metrology results at process time; this goal is usually achieved by means of statistical models, linking process ...
ABSTRACT In this paper, a multiple classifier machine learning (ML) methodology for predictive ma... more ABSTRACT In this paper, a multiple classifier machine learning (ML) methodology for predictive maintenance (PdM) is presented. PdM is a prominent strategy for dealing with maintenance issues given the increasing need to minimize downtime and associated costs. One of the challenges with PdM is generating the so-called “health factors,” or quantitative indicators, of the status of a system associated with a given maintenance issue, and determining their relationship to operating costs and failure risk. The proposed PdM methodology allows dynamical decision rules to be adopted for maintenance management, and can be used with high-dimensional and censored data problems. This is achieved by training multiple classification modules with different prediction horizons to provide different performance tradeoffs in terms of frequency of unexpected breaks and unexploited lifetime, and then employing this information in an operating cost-based maintenance decision system to minimize expected costs. The effectiveness of the methodology is demonstrated using a simulated example and a benchmark semiconductor manufacturing maintenance problem.
ABSTRACT Smart management of maintenances has become fundamental in manufacturing environments in... more ABSTRACT Smart management of maintenances has become fundamental in manufacturing environments in order to decrease downtime and costs associated with failures. Predictive Maintenance (PdM) systems based on Machine Learning (ML) techniques have the possibility with low added costs of drastically decrease failures-related expenses; given the increase of availability of data and capabilities of ML tools, PdM systems are becoming really popular, especially in semiconductor manufacturing. A PdM module based on Classification methods is presented here for the prediction of integral type faults that are related to machine usage and stress of equipment parts. The module has been applied to an important class of semiconductor processes, ion-implantation, for the prediction of ion-source tungsten filament breaks. The PdM has been tested on a real production dataset.
2011 IEEE International Conference on Automation Science and Engineering, 2011
The present paper is motivated by the application of Predictive Maintenance (PM) techniques in th... more The present paper is motivated by the application of Predictive Maintenance (PM) techniques in the semiconduc- tor manufacturing environment: such techniques are able, using process data, to make reliable predictions of residual equipment lifetime. The employment of PM yields positive fallouts on the productive process in form of unscheduled downtime reduction, increased spare parts availability and improved overall pro- duction
Uploads
Papers by Andrea Schirru