ENBIS-17 in Naples

9 – 14 September 2017; Naples (Italy) Abstract submission: 21 November 2016 – 10 May 2017

My abstracts

 

The following abstracts have been accepted for this event:

  • A Note on the Minimum Sample Size and Critical Acceptance Value for Variables Sampling Inspection Scheme Based on Process Capability Indices Cpk and Cpmk

    Authors: Antonio Lepore (University of Naples Federico II), Biagio Palumbo (University of Naples Federico II), Guido Cesaro (University of Naples Federico II)
    Primary area of focus / application: Quality
    Secondary area of focus / application: Six Sigma
    Keywords: Quality control, Acceptance sampling, Process capability indices, Fuel consumption monitoring, Decision-making
    Submitted at 10-May-2017 09:11 by Biagio Palumbo
    Accepted
    11-Sep-2017 17:50 A Note on the Minimum Sample Size and Critical Acceptance Value for Variables Sampling Inspection Scheme Based on Process Capability Indices Cpk and Cpmk
    Mathematical insights into the use of process capability indices (PCIs) Cpk and Cpmk are discussed in order to support managerial decision-making via hypothesis testing and avoid misleading conclusions [1]. In view of those arguments, inappropriate approximations found in the literature concerning producer’s α-risk and consumer’s β-risk corresponding to a sampling inspection scheme based on those PCIs are highlighted. In particular, in the case of Cpk, we demonstrate the approximation proposed in [2] gives required sample size and critical acceptance value that are not (even asymptotically) conservative with respect to the α-risk. Some counter examples illustrate that such approximation leads to substantial deviations from the actual α-risk and thus, should be not used in practice. Moreover, we point out that there is no need of particular computer programs and tables to numerically determine the exact solution. Nevertheless, the same issue occurs in [3] and a similar yet less severe inaccuracy with respect to the case of the Cpmk is also noticed in [4].

    REFERENCES
    [1] Lepore, A., & Palumbo, B. (2015). New Insights into the Decisional Use of Process Capability Indices via Hypothesis Testing. Quality and Reliability Engineering International, 31(8), 1725–1741. https://doi.org/10.1002/qre.1713
    [2] Wu, C. W., Aslam, M., & Jun, C. H. (2012). Variables sampling inspection scheme for resubmitted lots based on the process capability index Cpk. European Journal of Operational Research, 217(3), 560–566. https://doi.org/10.1016/j.ejor.2011.09.042
    [3] Pearn, W. L., & Wu, C. W. (2007). An effective decision making method for product acceptance. Omega, 35(1), 12–21. https://doi.org/10.1016/j.omega.2005.01.018
    [4] Wu, C. W., & Pearn, W. L. (2008). A variables sampling plan based on Cpmk for product acceptance determination. European Journal of Operational Research, 184(2), 549–560. https://doi.org/10.1016/j.ejor.2006.11.032
  • Misdiagnosis in Emergency Department and Its Impact on Hospital Load

    Authors: Yariv Marmor (ORT Braude College of Engineering)
    Primary area of focus / application: Other: Process improvement in healthcare
    Keywords: Emergency department, Diagnosis, Misclassification, Tree structured data, Performance
    Submitted at 10-May-2017 10:02 by Yariv Marmor
    Accepted (view paper)
    13-Sep-2017 09:20 Misdiagnosis in Emergency Department and Its Impact on Hospital Load
    Accurate diagnosis of patients is one of emergency departments (ED) main objectives, where patient health state can be characterized by a single category or by a set of multi health issues. In our work, we adopted the International Statistical Classification of Diseases and Related Health Problems (ICD) codes in its 9th version. We, first, suggest metric for measuring diagnosis divergence between two major diagnoses – one that is made to finalize ED process and the second that concludes patient’s hospitalization. As the ICD codes are hierarchical based, there is a need to account for appropriate measures. Next, we propose a way to compare between patient’s comorbidity (multiple health conditions) after ED process and at the end of hospitalization based on new approach of ANOVA for tree-structured data. Finally, we explore a case study in order to demonstrate how the load at the ED is related to an effective classification, and how misclassification is related to an extra work in the hospital wards.
  • Knowledge Improvement of an Additive Manufacturing Process via Stochastic Approach: Applicative Examples

    Authors: Biagio Palumbo (University of Naples Federico II), Francesco Del Re (University of Naples Federico II), Pasquale Corrado (MBDA ITALIA), Giuseppe La Sala (MBDA ITALIA)
    Primary area of focus / application: Design and analysis of experiments
    Secondary area of focus / application: Quality
    Keywords: Additive manufacturing, Direct metal laser sintering, Knowledge improvement, Stochastic approach, Design of Experiments
    Submitted at 10-May-2017 12:08 by Francesco Del Re
    Accepted
    12-Sep-2017 10:30 Knowledge Improvement of an Additive Manufacturing Process via Stochastic Approach: Applicative Examples
    Additive Manufacturing, also known as 3D printing, is considered one of the most attractive key enabling technologies in the framework of the so-called Industry 4.0, that is the new paradigm of the Fourth Industrial Revolution.
    The aim of this work is to highlight, through three applicative examples, the strategic role that a systematic stochastic approach can play in the knowledge improvement of an Additive Manufacturing process based on the Direct Metal Laser Sintering technology, where three-dimensional components are produced from metal powder.
    In particular, the following key technological issues have been investigated: the effects of metal powder reusing and of topological printing strategies (i.e. component position and orientation within the printing volume) on the mechanical properties as well as the optimal settings of laser exposure parameters in terms of process productivity.
    The experimental activities have been carried out after an accurate pre-design phase in order to reduce efforts, costs and time required to investigate the process.
    The applicative examples have been developed in partnership with MBDA ITALIA company and represent a positive example of synergic integration between statistical and technological competences.
  • Distributions for the Maximum of a Random Number of Observations

    Authors: Marta Perez-Casany (Universitat Politècnica de Catalunya)
    Primary area of focus / application: Reliability
    Secondary area of focus / application: Modelling
    Keywords: Extreme value theory, Generalized extreme value distribution, Generalized Pareto distribution, Positive Poisson distribution
    Submitted at 10-May-2017 12:56 by Marta Perez-Casany
    Accepted
    11-Sep-2017 16:40 Distributions for the Maximum of a Random Number of Observations
    Extreme values theory is the branch of statistics that deal with the problem of modelling and predicting events that are far away from the its usual range. It is required in many research areas as industry (for predicting important system malfunctions), in insurance ( for predicting huge number of claims due to catastrophic events), in environmental studies ( for predicting excess of contamination in big cities), etc. This theory studies the distribution of maximums, events that are much larger than the usual observations, and its main theorem, known as Fisher-Tippet-Gnedenko theorem, establishes that the maximum follows asymptotically a Generalized Extreme Value (GEV) distribution, but the convergence is, in general, slow. Another important theorem is the one due to Pickards, Balkema and de Haan that states that the observations over a threshold have as asymptotic distribution a Generalized Pareto distribution (GP).

    In many practical cases, one is interested in the maximum (minimum) value of a given measure during a given period of time or space, without knowing in advance how many observations there will be. In that case one needs to consider maximums (minimums) of a random number of independent copies of a given random variable, which is denoted as random stopped extreme distributions (RSED). The RSED depends on the distribution of the phenomena and on the distribution of the number of events. The main advantage of using RSED is that it does not require to have a large number of observations. The classical theory of Extreme values may be seen as a particular case. Emphasis will be given to the particular case where the number of observations follows a Positive Poisson distribution.
  • A Stochastic Approach for Aerospace Industry Innovation via Engineering and Statistical Knowledge Integration

    Authors: Gaetano De Chiara (Avio Aero), Roberto Marrone (Avio Aero), Biagio Palumbo (University of Naples, Federico II)
    Primary area of focus / application: Other: Quality engineering applied to advanced manufacturing
    Keywords: Six-Sigma, Laser drilling, Pre-experimental planning phase, Design of Experiment
    Submitted at 10-May-2017 13:09 by Roberto Marrone
    Accepted
    11-Sep-2017 11:10 A Stochastic Approach for Aerospace Industry Innovation via Engineering and Statistical Knowledge Integration
    Systematic and sequential approaches to industrial experimentation are known to be crucial in the process innovation and continuous improvement of aerospace industry processes. The integration between engineering knowledge and statistical thinking allows for a virtuous cycle of sequential learning consistent with Six Sigma approach.
    In this paper, the effectiveness of such approaches is demonstrated through several examples of laser drilling of combustion chambers for aerospace industry. In this technological field, many thousands of high quality small and shaped effusion holes need to be realized. The proposed examples have been developed within the Avio Aero industry, which is an aerospace company at the leading edge of propulsion technology. They show the continuous technological innovation of laser process in Avio Aero from conventional to fiber laser sources.
    All experimental activities have been executed adopting a systematic approach in the pre-experimental planning phase where pre-design sheets have been conceived and implemented ad hoc. The remarkable results obtained in the last years make the Avio Aero plant in Pomigliano d’Arco the only European alternative for the production of advanced combustion chambers to GE Aviation American sites.
  • Prediction of Coating Properties with Generalized Linear Models in an HVOF Spraying Process

    Authors: Eva-Christina Becker-Emden (FH Dortmund), Sonja Kuhnt (FH Dortmund)
    Primary area of focus / application: Modelling
    Keywords: HVOF spraying process, Generalized linear models, Model selection, Prediction of coating properties, Different spraying powders
    Submitted at 10-May-2017 14:01 by Eva-Christina Becker-Emden
    Accepted
    11-Sep-2017 16:00 Prediction of Coating Properties with Generalized Linear Models in an HVOF Spraying Process
    High velocity oxygen fuel (HVOF) spraying is a method to apply protective coatings on surfaces. Obtaining properties of such coatings is expensive and time consuming, hence it is desirable to derive predictions based on parameters or particle properties that are easier to measure. It is also of interest to investigate different powders used in the industry.

    We use generalized linear models with different link functions and gamma distributed responses to describe the coating properties in dependence of particle properties, like particle temperature or velocity, and process parameters, like the feeder disc velocity. Previously only a WC-12Co powder of type WOKA 3102 from Sulzer Metco was analysed. In this contribution, we run experiments on two further powders, namely WC-FeCrAl and WC-Co-100nm. We analyse to which extend the models for WC-12Co are transferable to other powders. Additionally, individual model selection for the new powders is conducted and results are compared. We base our model building on an all-subset selection with the BIC as selection criterion, and a maximal model containing main, two-way interaction and quadratic effects.

    We find that the class of generalized linear models provides adequate prediction models for the individual coating properties. Possible extensions of the individual generalized linear models to a multivariate model are discussed.