Ioan C. Bacivarov

Control charts for non-Gaussian distributions

Traditional statistical process control (SPC) techniques applied in the industrial processes field consider often that the distribution ofdata is Gaussian. The estimation ofparameters, the detection ofthe out oforder situations and the control of the followed characteristics are easy to achieve for the normal populations. In reality, whatever the origin of a characteristic (large series productions for components, mechanical parts of OE communication systems, etc. ) the curve of distributions of the measured values is generally far from being normal. The simple approximation to the Gauss distribution and the use of the classical control methods sometimes induces serious errors. In this paper, a study on the statistical control of non Gaussian populations is presented. Particularly we discuss the Rayleigh and the Weibull distribution as being representatives in (SPC for some category of data. The X control charts with variable limits are tested. Experimental simulations are presented for different parameters of the two distributions. The results confirm the methodology and encourage the research in the field of non Gaussian processes.

Studying cyber security threats to web platforms using attack tree diagrams

Used by both information systems designers and security personnel, the Attack Tree method provides a graphical analysis of the ways in which an entity (a computer system or network, an entire organization, etc.) can be attacked and indicates the countermeasures that can be taken to prevent the attackers to reach their objective. In this paper, we built an Attack Tree focused on the goal “compromising the security of a Web platform”, considering the most common vulnerabilities of the WordPress platform identified by CVE (Common Vulnerabilities and Exposures), a global reference system for recording information regarding computer security threats. Finally, based on the likelihood of the attacks, we made a quantitative analysis of the probability that the security of the Web platform can be compromised.

Intruder monitoring system for local networks using Python

Nowadays, Information Security specialists focus on the development of complex solutions for monitoring large enterprise networks. This paper is intended to document a security solution developed within Python programming language created for the use in LANs. Based on Open-Source software and constructed to run on low-cost hardware, this program is designed to run ad-hoc or on demand scanning across a network and notify the owner through alarms in case of intrusions. This article will additionally analyze the programs integration with related concepts like CIA (Confidentiality, Integrity, and Availability) of Information Security, Defense in Depth Model and Cyber Kill Chain.

E-learning platform for automated testing of electronic circuits using signature analysis method

Dependability of electronic circuits can be ensured only through testing of circuit modules. This is done by generating test vectors and their application to the circuit. Testability should be viewed as a concerted effort to ensure maximum efficiency throughout the product life cycle, from conception and design stage, through production to repairs during products operating. In this paper, is presented the platform developed by authors for training for testability in electronics, in general and in using signature analysis method, in particular. The platform allows highlighting the two approaches in the field namely analog and digital signature of circuits. As a part of this e-learning platform, it has been developed a database for signatures of different electronic components meant to put into the spotlight different techniques implying fault detection, and from this there were also self-repairing techniques of the systems with this kind of components. An approach for realizing self-testing circuits based on MATLAB environment and using signature analysis method is proposed. This paper analyses the benefits of signature analysis method and simulates signature analyzer performance based on the use of pseudo-random sequences, too.

Quality politics: an immaterial investment for companies in (micro)electronics

With the globalization of the markets and the growth of competitiveness in the manufacturing sector, quality has become a key factor of success. Quality is particularly important for the companies which activate in the micro(electronics) field. The quality management system holds a vital place in the companys structure. Implementing such a system requires important operating costs. These costs are known as Quality Obtaining Costs (QOC) and may be considered as an investment. Planning an investment, means evaluating its return in order to see if it is profitable or not. Measuring the return of quality politics investment raise some delicate problems. We may calculate some aspects of the return of investment by measuring the shape of non-quality costs. An eventual decrease of these costs could be synonym with a profitable investment. But the advantages of good quality politics cannot be measured only by taking into consideration the non-quality costs (even if they include direct and indirect costs). There are also intangible advantages (like mark image, competences, polyvalence, clients satisfaction...) that derive from quality approaches. How to evaluate this type of consequences / advantages? The idea developed in this article is to considerate the quality politics like un immaterial/intelligent investment. Therefore could it be advantageous / possible to use the immaterial investments measuring and evaluation techniques for studying the quality politics return of investment?

A method for reliability estimation of heterogeneous systems

Reliability estimation is becoming an important issue of the design process of complex heterogeneous systems. The concept of reliability is frequently seen as being one of the least controlled points and for some as being the critical point. Since these systems are very complex to study, the evaluation of their reliability is extremely difficult. In this paper, we propose a global method to estimate the mechatronic system reliability using operating field data. Since we have a small amount of data, we use an estimation method called Bayesian Restoration Maximization (BRM) method, thus increasing the estimation accuracy. The BRM method needs to define some prior knowledge. For this purpose, we propose to define the prior distribution using a Monte-Carlo simulation based on stochastic Petri Nets (SPN) model and on the operating field data. The stochastic PN model describes the functional and dysfunctional behaviours. In this study, we deal with the case of n repairable systems until a deterministic censoring time (for example, this censoring time may be the warranty period of an ABS system). We consider repair as the replacement of the failing component by an identical one in the case of electronic and mechanical subsystem and in the case of software, the default is rectified on all the subsystems. We simulate the failures times and we compute the confidence interval. The proposed method allows reliability evaluating both for n mechatronic systems and for their different subsystems.

Quality assurance and management in microelectronics companies: ISO 9000 versus Six Sigma

A strategy for the implementation of the Six Sigma method as an improvement solution for the ISO 9000:2000 Quality Standard is proposed. Our approach is focused on integrating the DMAIC cycle of the Six Sigma method with the PDCA process approach, highly recommended by the standard ISO 9000:2000. The Six Sigma steps applied to each part of the PDCA cycle are presented in detail, giving some tools and training examples. Based on this analysis the authors conclude that applying Six Sigma philosophy to the Quality Standard implementation process is the best way to achieve the optimal results in quality progress and therefore in customers satisfaction.

Life testing of electronic components. A Bayesian approach

A common problem of high reliability computing is, on one hand, the magnitude of total testing time required, particularly in the case of high reliability components and, on the other hand, the number of devices under test. In both cases, the objective is to minimize the costs involved in testing without reducing the quality of the data obtained. One solution is based on accelerated life testing techniques which permit to decrease testing time. Another solution is to incorporate prior beliefs, engineering experience, or previous data into the testing framework. It is in this spirit that the use of a Bayesian approach can, in many cases, significantly reduce the amount of devices required. The accelerated life testing (ALT) of electronic components (including the semiconductor devices) under severer than operating conditions involving high temperature, humidity, voltage, a.o. is commonly used to reduce test time and cost. The main problem is to estimate accelerated life model parameters allowing to define the reliability function under operating conditions from only accelerated life data. A difficulty of using the ALT, during design stage, is due to the small sample size to test. In this context, the Bayesian approach can be used to incorporate into the estimation process all available knowledge on accelerated life model (baseline failure rate, activation energy, a.o). This paper presents the study of Exponential-Arrhenius model by an evaluation of parameters using maximum likelihood and Bayesian methods. A Monte Carlo simulation has been performed to examine the asymptotic behavior of these different estimators.