JOURNAL OF INDUSTRIAL ENGINEERING INTERNATIONAL
Year:2018 | Volume:14 | Issue:2|Papers Count:15

This paper first enlists the generic problems ofalloy wheel machining and subsequently details on theprocess improvement of the identified critical-to-qualitymachining characteristic of A356 aluminum alloy wheelmachining process. The causal factors are traced using theIshikawa diagram and prioritization of corrective actions isdone through process failure modes and effects analysis. Process monitoring charts are employed for improving theprocess capability index of the process, at the industrialbenchmark of four sigma level, which is equal to the valueof 1. 33. The procedure adopted for improving the processcapability levels is the define-measure-analyze-improvecontrol(DMAIC) approach. By following the DMAICapproach, the Cp, Cpk and Cpm showed signs of improvementfrom an initial value of 0. 66,-0. 24 and 0. 27, to afinal value of 4. 19, 3. 24 and 1. 41, respectively.

Furniture manufacturing micro and small enterprisesare confronted with several factors that affect theirperformance. Some enterprises fail to sustain, some othersremain for long period of time without transforming, andmost are producing similar and non-standard products. Themain aim of this manuscript is on improving the performanceand contribution of MSEs by analyzing impact ofsignificant internal and external factors. Data was collectedvia a questionnaire, group discussion with experts andinterviewing process. Randomly selected eight representativemain cities of Amhara region with 120 furnituremanufacturing enterprises are considered. Data analysisand presentation was made using SPSS tools (correlation, proximity, and T test) and impact-effort analysis matrixtool. The correlation analysis shows that politico-legal withinfrastructure, leadership with entrepreneurship skills andfinance and credit with marketing factors are those factors, which result in high correlation with Pearson correlationvalues of r = 0. 988, 0. 983, and 0. 939, respectively. Thestudy investigates that the most critical factors faced byMSEs are work premises, access to finance, infrastructure, entrepreneurship and business managerial problems. Theimpact of these factors is found to be high and is confirmedby the 50% drop-out rate in 2014/2015. Furthermore, morethan 25% work time losses due to power interruption dailyand around 65% work premises problems challengedMSEs. Further, an impact-effort matrix was developed tohelp the MSEs to prioritize the affecting factors.

Cooperative and gradual covering are two newmethods for developing covering location models. In thispaper, a cooperative maximal covering location– allocationmodel is developed (CMCLAP). In addition, both cooperativeand gradual covering concepts are applied to themaximal covering location simultaneously (CGMCLP). Then, we develop an integrated form of a cooperativegradual maximal covering location problem, which iscalled a general CGMCLP. By setting the model parameters, the proposed general model can easily be transformedinto other existing models, facilitating general comparisons. The proposed models are developed without allocationfor physical signals and with allocation for nonphysicalsignals in discrete location space. Comparison ofthe previously introduced gradual maximal covering locationproblem (GMCLP) and cooperative maximal coveringlocation problem (CMCLP) models with our proposedCGMCLP model in similar data sets shows that the proposedmodel can cover more demands and acts more efficiently. Sensitivity analyses are performed to show theeffect of related parameters and the model’ s validity. Simulated annealing (SA) and a tabu search (TS) are proposedas solution algorithms for the developed models forlarge-sized instances. The results show that the proposedalgorithms are efficient solution approaches, consideringsolution quality and running time.

Tri-level optimization problems are optimizationproblems with three nested hierarchical structures, where inmost cases conflicting objectives are set at each level ofhierarchy. Such problems are common in management, engineering designs and in decision making situations ingeneral, and are known to be strongly NP-hard. Existingsolution methods lack universality in solving these types ofproblems. In this paper, we investigate a tri-level programmingproblem with quadratic fractional objectivefunctions at each of the three levels. A solution algorithmhas been proposed by applying fuzzy goal programmingapproach and by reformulating the fractional constraints toequivalent but non-fractional non-linear constraints. Basedon the transformed formulation, an iterative procedure isdeveloped that can yield a satisfactory solution to the trilevelproblem. The numerical results on various illustrativeexamples demonstrated that the proposed algorithm is verymuch promising and it can also be used to solve largersizedas well as n-level problems of similar structure.

This paper proposes a new motivation forinformation sharing in a decentralized channel consistingof a single manufacturer and two competing retailers. Themanufacturer provides a common product to the retailers atthe same wholesale price. Both retailers add their ownvalues to the product and distribute it to consumers. Factorssuch as retail prices, values added to the product, and localadvertising of the retailers simultaneously have effect onmarket demand. Each retailer has full information aboutthe own added value which is unknown to the manufacturerand other retailer. The manufacturer uses a cooperativeadvertising program for motivating the retailers to disclosetheir private information. A numerical study is presented tocompare different scenarios of information sharing. Computationalresults show that there is a condition in whichfull information sharing is beneficial for all members of thesupply chain through cooperative advertising program and, therefore, retailers have enough incentive to disclose theircost information to the manufacturer.

In many cases, redundant systems are beset byboth independent and dependent failures. Ignoring dependentvariables in MTBF evaluation of redundant systemshastens the occurrence of failure, causing it to take placebefore the expected time, hence decreasing safety andcreating irreversible damages. Common cause failure(CCF) and cascading failure are two varieties of dependentfailures, both leading to a considerable decrease in theMTBF of redundant systems. In this paper, the alpha-factormodel and the capacity flow model are combined so as toincorporate CCF and cascading failure in the evaluation ofMTBF of a 2-out-of-3 repairable redundant system. Then, using a transposed matrix, the MTBF function of the systemis determined. Due to the fact that it is difficult toestimate the independent and dependent failure rates, industries are interested in considering uncertain failurerates. Therefore, fuzzy theory is used to incorporateuncertainty into the model presented in this study, and anonlinear programming model is used to determine system’ sMTBF. Finally, in order to validate the proposedmodel, evaluation of MTBF of the redundant system of acentrifugal water pumping system is presented as a practicalexample.

For this paper, we explored the implementationof the cuckoo search algorithm applied to the capacitatedvehicle routing problem. The cuckoo search algorithm wasimplemented with Le´ vy flights with the 2-opt and doublebridgeoperations, and with 500 iterations for each run. Thealgorithm was tested on the problem instances from theAugerat benchmark dataset. The algorithm did not performwell on the problem instances, save for a select few onwhich the algorithm achieved the close to near-optimalresult and one on which the algorithm achieved the optimalresult. Increasing the number of iterations for each run ofthe algorithm on the two large-scale problem instances ledto obtaining solutions closer to the optimal solution comparedto the ones obtained with fewer number iterations. This gives an idea that the larger the problem instancebecomes, the slower the algorithm converges to the optimalsolution. Several other factors may also have contributed tothe overall performance of the algorithm. Regardless of itsperformance, the algorithm was able to obtain routes thatsatisfied the constraints of the capacitated vehicle routingproblem. The potential of the cuckoo search algorithm insolving combinatorial problems is demonstrated in thisstudy in which the performance of the algorithm on routingproblems was explored.

This study aims to discuss the solutionmethodology for a closed-loop supply chain (CLSC) networkthat includes the collection of used products as wellas distribution of the new products. This supply chain ispresented on behalf of the problems that can be solved bythe proposed meta-heuristic algorithms. A mathematicalmodel is designed for a CLSC that involves three objectivefunctions of maximizing the profit, minimizing the totalrisk and shortages of products. Since three objectivefunctions are considered, a multi-objective solutionmethodology can be advantageous. Therefore, severalapproaches have been studied and an NSGA-II algorithm isfirst utilized, and then the results are validated using anMOSA and MOPSO algorithms. Priority-based encoding, which is used in all the algorithms, is the core of thesolution computations. To compare the performance of themeta-heuristics, random numerical instances are evaluatedby four criteria involving mean ideal distance, spread ofnon-dominance solution, the number of Pareto solutions, and CPU time. In order to enhance the performance of thealgorithms, Taguchi method is used for parameter tuning. Finally, sensitivity analyses are performed and the computationalresults are presented based on the sensitivityanalyses in parameter tuning.

Nowadays, organizations have to compete withdifferent competitors in regional, national and internationallevels, so they have to improve their competition capabilitiesto survive against competitors. Undertaking activitieson a global scale requires a proper distribution systemwhich could take advantages of different transportationmodes. Accordingly, the present paper addresses a location-routing problem on multimodal transportation network. The introduced problem follows four objectivessimultaneously which form main contribution of the paper; determining multimodal routes between supplier and distributioncenters, locating mode changing facilities, locatingdistribution centers, and determining product deliverytours from the distribution centers to retailers. An integerlinear programming is presented for the problem, and agenetic algorithm with a new chromosome structure proposedto solve the problem. Proposed chromosome structureconsists of two different parts for multimodaltransportation and location-routing parts of the model. Based on published data in the literature, two numericalcases with different sizes generated and solved. Also, differentcost scenarios designed to better analyze model andalgorithm performance. Results show that algorithm caneffectively solve large-size problems within a reasonabletime which GAMS software failed to reach an optimalsolution even within much longer times.

A deterministic inventory model with two levelsof storage (own warehouse and rented warehouse) withnon-instantaneous deteriorating items is studied. The supplieroffers the retailer a trade credit period to settle theamount. Different scenarios based on the deterioration andthe trade credit period have been considered. In this article, we have framed two models considering single warehouse(Model-I) and two warehouses (Model-II) for non-instantaneousdeteriorating items. The objective of this work is tominimize the total inventory cost and to find the optimallength of replenishment and the optimal order quantity. Mathematical theorems have been developed to determinethe existence and the uniqueness of the optimal solution. Computational algorithms for the two different models aredesigned to find the optimal order quantity and the optimalcycle time. Comparison between the optimal solutions forthe two models is also given. Numerical illustrations andmanagerial insights obtained demonstrate the applicationand the performance of the proposed theory.

Decentralized supply chain management isfound to be significantly relevant in today’ s competitivemarkets. Production and distribution planning is posed asan important optimization problem in supply chain networks. Here, we propose a multi-period decentralizedsupply chain network model with uncertainty. The imprecisionrelated to uncertain parameters like demand andprice of the final product is appropriated with stochasticand fuzzy numbers. We provide mathematical formulationof the problem as a bi-level mixed integer linear programmingmodel. Due to problem’ s convolution, a structureto solve is developed that incorporates a novelheuristic algorithm based on Kth-best algorithm, fuzzyapproach and chance constraint approach. Ultimately, anumerical example is constructed and worked through todemonstrate applicability of the optimization model. Asensitivity analysis is also made.

The choice of suitable robots in manufacturing, to improve product quality and to increase productivity, isa complicated decision due to the increase in robot manufacturersand configurations. In this article, a novelapproach is proposed to choose among alternatives, differentlyassessed by decision makers on different criteria, to make the final evaluation for decision-making. Theapproach is based on the ellipsoid algorithm for systems oflinear inequalities. Most of the ranking methods depend onintegration that becomes complicated for nonlinear membershipfunctions, which is the case in robot selection. Themethod simply uses the membership function or itsderivative. It takes the decision maker’ s attitude in ranking. It effectively ranks fuzzy numbers and their images, preservingsymmetry. It is a simple recursive algebraic formulathat can be easily programmed. The performance ofthe algorithm is compared with the performance of someexisting methods through several numerical examples toillustrate its advantages in ranking, and a robot selectionproblem is solved.

This paper questions some aspects of attributeacceptance sampling in light of the original concepts ofhypothesis testing from Neyman and Pearson (NP). Attributeacceptance sampling in industry, as developed byDodge and Romig (DR), generally follows the internationalstandards of ISO 2859, and similarly the Brazilian standardsNBR 5425 to NBR 5427 and the United StatesStandards ANSI/ASQC Z1. 4. The paper evaluates andextends the area of acceptance sampling in two directions. First, by suggesting the use of the hypergeometric distributionto calculate the parameters of sampling plansavoiding the unnecessary use of approximations such as thebinomial or Poisson distributions. We show that, underusual conditions, discrepancies can be large. The conclusionis that the hypergeometric distribution, ubiquitouslyavailable in commonly used software, is more appropriatethan other distributions for acceptance sampling. Second, and more importantly, we elaborate the theory of acceptancesampling in terms of hypothesis testing rigorouslyfollowing the original concepts of NP. By offering acommon theoretical structure, hypothesis testing from NPcan produce a better understanding of applications evenbeyond the usual areas of industry and commerce such aspublic health and political polling. With the new procedures, both sample size and sample error can be reduced. What is unclear in traditional acceptance sampling is thenecessity of linking the acceptable quality limit (AQL)exclusively to the producer and the lot quality percentdefective (LTPD) exclusively to the consumer. In reality, the consumer should also be preoccupied with a value ofAQL, as should the producer with LTPD. Furthermore, wecan also question why type I error is always uniquelyassociated with the producer as producer risk, and likewise, the same question arises with consumer risk which isnecessarily associated with type II error. The resolution ofthese questions is new to the literature. The article presentsR code throughout.

The need of taxonomy is vital for knowledge sharing. This need has been portrayed by through-life engineering services/systems. This paper addresses this issue by repair process taxonomy development. Framework for repair process taxonomywas developed followed by its implementation. The importance of repair process taxonomy has been highlighted.

This paper aims at proposing a quadratic assignment-based mathematical model to deal with the stochasticdynamic facility layout problem. In this problem, productdemands are assumed to be dependent normally distributedrandom variables with known probability density functionand covariance that change from period to period at random. To solve the proposed model, a novel hybrid intelligentalgorithm is proposed by combining the simulated annealingand clonal selection algorithms. The proposed model and thehybrid algorithm are verified and validated using design ofexperiment and benchmark methods. The results show thatthe hybrid algorithm has an outstanding performance fromboth solution quality and computational time points of view. Besides, the proposed model can be used in both of thestochastic and deterministic situations.