w@reRISK method: Security risk level classification of stock keeping units in a warehouse

Safety Science 79 (2015) 358–368

Contents lists available at ScienceDirect

Safety Science

journal homepage: www.elsevier .com/locate /ssc i

w@reRISK method: Security risk level classification of stock keepingunits in a warehouse

E-mail address: gaston.cedillo@mexico-logistico.org (M.G. Cedillo-Campos)

Miguel Gastón Cedillo-Campos a, Hermes Orestes Cedillo-Campos b

a Mexican Institute of Transportation, Transportation and Logistics Systems National Laboratory, Carretera Querétaro-Galindo Km 12, Sanfandila,Mpio. Pedro Escobedo, C.P. 76703 Queretaro, Mexicob YoLogistico.COM, Francisco P. Mariel No. 155, C.P. 78233 San Luis Potosí, Mexico

a r t i c l e i n f o

Article history:Received 22 January 2015Received in revised form 8 May 2015Accepted 11 June 2015

Keywords:Supply chain securityRisk managementWarehouse managementInventory systemsMulticriteria analysis

a b s t r a c t

Risk management in warehouses is a key issue to guarantee security all along the global supply chains.However, from a balanced approach including at the same time a solid theoretical background and prac-tical processes of implementation, our comprehensive literature and field practices review proved thatmost of current methods to manage warehouses still do not take into account the risk level classificationof stock keeping units (SKU). Thus, the w@reRISK method to analyze the security risk of SKU stocked in awarehouse is here exposed. Since risk analysis involves not only factual data, but also perceptions, thispaper proposes a hybrid method based on the ABC classification of SKU as well as a XYZ variability systemanalysis, but also incorporates the Analytic Hierarchy Process (AHP) to ponder subjective values whichare part of risk. The method includes a security risk level classification of SKU, then introduces an exhaus-tive list of 34 factors useful to judge the risk of a SKU stocked in a warehouse for comparison purposes. Anapplication example is included to demonstrate the use of the method. The lack of analytical frameworkswith solid technical basis to secure warehouses along the global supply chains makes this paper acontribution to the body of knowledge in supply chain security, as well as a friendly-user method topractitioners. The proposed method is susceptible to become an automated informatics tool. As aconclusion, key issues for designing secure manufacturing supply chains are discussed, and futureresearch is presented.

1. Introduction

Supply chain security is a multidimensional area of knowledge,which should be engaged from a comprehensive as well as anintegrated point of view. Since a large proportion of the cargo iscompletely or partially loaded through different warehouses, theselogistics facilities become a critical link inside supply chains.Assessing all the combination of risks related to different stockkeeping units (SKU) in a warehouse is complex. Even if someSKU have an evident level of risk which allows decision makersto define the right location where the SKU should be placed in awarehouse (computers, flash memories, cell phones, which arehighly marketable), others are less obvious. This is the case of cer-tain inoffensive (when used in everyday life) chemical or informat-ics products, which could be combined or assembled to produce apotential weapon. To place these kinds of products close to them ina warehouse could not only cause damage by accident, but also

enables their illegal use by criminals. Since a specific risk analysisof every product in each warehouse along a supply chain istime-consuming and requires highly trained personnel, an inte-grated method based on a standardized approach is needed.

Thus, in the U.S., from a criminal risk perspective, the NBCrecently reported that more than $1.2 million worth of computers,cameras and other electronic devices were stolen from warehousesin Fremont, California (NBC, 2014). From a simplistic point of view,it is a security issue which could be understood as a result of aweak policing or company’s security systems. However, in emerg-ing markets as Mexico, the multidimensional nature of securityrisk to supply chains is more evident. Since product type cargothefts vary between regions, security risk level classification ofSKU is not trivial. Electronics thefts frequently occur close tometropolitan areas in the central region of the country, while steeland other metals are more targeted in the north. High-value itemsand food remain the most pursued after cargo by thieves, as theyare readily sold or marketable (U.S. Overseas Security AdvisoryCouncil, 2014). On the other hand, from a national security

W@reRISK

ABCanalysis

XYZanalysis

AHPrisk

analysis

Fig. 1. w@reRISK method.

M.G. Cedillo-Campos, H.O. Cedillo-Campos / Safety Science 79 (2015) 358–368 359

perspective and according to Bakir and Pakdaman (2006):‘‘US-bound containerized overseas cargo move through variousphases, all of which present unique security challenges. Issues in con-tainer security can be summarized under five phases: loading phase atthe warehouse, land transportation, port of origin, sea transportationand port of destination’’. Thus, Bakir and Pakdaman (2006) high-lighted that terrorists may load illegal weapons and explosives atthe warehouses or distribution centers from which the cargo isdispatched. In that context, security breaches can sometimes com-promise the existence of the companies (Markert, 1998; Johansson,2008; Torabi et al., 2014).

Consequently, when analyzing the true cost of supply chaincrime, Jones et al. (2009) identified an ‘‘Iceberg Effect’’. Their analy-sis reported $15–18 billion in direct costs, but 60 billion of indirectcosts in losses (sales, reputation, insurances, investigation cost,re-order costs, administrative costs, paying claims, and others).

For decades, the constant quest for efficiency in productiveoperations has led companies to pursue the zero inventory andjust-in-time paradigm through philosophies such as ‘‘LeanManufacturing’’. Yet, the increase in demand variability and the risklinked to security threats on global supply chains, are reshapingthe way to design, organize and manage more resilient logisticssystems (Cedillo-Campos and Perez, 2010; GAO, 2010; Pfohlet al., 2010; Thun and Hoening, 2011; Vilko and Hallikas, 2011;The White House, 2012; Elmoselhy, 2014; Cedillo-Campos et al.,2014). In fact, modern companies are now looking for improvingtheir ‘‘supply chain fluidity’’, which is understood as the capabilitydegree to achieve continuously a reliable, secure, and accurate flowof process, supporting effectively the supply chain goals. Given theimportant interrelation among actors, any disruptive effect, when asecurity risk materializes in the supply chain, can bring importanteconomic consequences for a company, as well as for other supplychain members as a result of risk propagation (Johansson, 2008;Bakshi and Gans, 2010; Serdarasan, 2013; Bueno andCedillo-Campos, 2014). Globalization has facilitated trade; how-ever, the supply chains are every day more vulnerable (Donnerand Kruk, 2009; Ochoa, 2010; Sodhi and Son, 2012; Kern et al.,2012; Cedillo-Campos, 2012; Chopra and Meindl, 2013); the com-plexity of current supply chains makes them highly vulnerable todirect impacts such as thefts, shrinkage or indirect impacts arisingfrom fines or other situations, as a result of non-fulfilling withstandards like C-TPAT (Customs-Trade Partnership AgainstTerrorism).

Since supply chain vulnerability is increasing, interest to under-stand risk and how to deal with the effects of the disruptions insupply chains has arisen as a global priority (Lee and Whang,2005; Buesa et al., 2007; Ochoa, 2010; Bueno andCedillo-Campos, 2014). Nowadays, threats are an internationalmatter that use supply chains of companies respectful of the lawto do harm. Containers, products, and modes of transportationcreate a screen, preventing quick detection of harmful contents,making them very attractive to criminals. In that sense, the idealwould be to implement 100% scanning policies at all points ofrupture of the cargo, but it would be virtually impossible (GAO,2008; Xue and Villalobos, 2012; Cedillo-Campos et al., 2014).

Actually, total security supply chain can only be reached if everylink, from the point of origin to its final destination, is responsiblefor the security of its part of the supply chain (Holler and Schanck,2010; Schlegel and Trent, 2012; The White House, 2012). Sincewarehouses are important, because of their number inside a supplychain and of their important role in the synchronization of manu-facturing flows, controlling security in warehouses makes a criticallink in the security of the whole supply chain.

In a modern and global supply chain, any warehouse must guar-antee the security of the products and of the workforce handlingthem. The warehouse is the link that must guarantee the

appropriate handling and custody of SKU. Depending on the typesof products located in the warehouses, different criteria are used totheir design, such as the classification of SKU, inventory turnover,environment conditions, and other specific aspects. Nevertheless,the lack of internationally standardized systems for warehousemanagement to reduce security risks makes them a highly vulner-able link in the supply chain.

Based on the global studies developed by Johansson (2008) andHintsa et al. (2010) and completed by a comprehensive literatureand field practices review, we concluded that, from a balancedapproach including at the same time a solid theoretical back-ground and a practical processes of implementation, methodolo-gies to improve warehouse management considering the securityrisk level of SKU were not found. Consequently, the objective ofthis paper is to develop an analytical framework for providingmanagerial insights for those dealing with security risk classifica-tion of SKU located in a warehouse.

The article is organized as follows: Section 2 exposes the analyt-ical framework as well as the opportunity area; Section 3 presentsan illustrative example, and finally, Section 4 discusses the con-cluding remarks and the future work derived from this work.

2. Analytical framework

For companies, the aim of reducing security breaches is to mit-igate the economic losses that may be due to the contamination orloss of products causing financial impacts as well as to keep safethe trademark prestige (Johansson, 2008; Hintsa et al., 2010;Bueno and Cedillo-Campos, 2014; Cedillo-Campos et al., 2014).However, in spite of the great number of international securityinitiatives, warehouse systems lack a practical, technicallyrobust, and not time-expensive method that enables evaluating,from a supply chain management perspective, different risks inwarehouses.

In order to cover this significant gap, our flexible and hybridmethod enables classifying products in a warehouse, using thecombination of different analytical tools taking into account: (i)Characteristics of how SKU are used (ABC); (ii) Variability; analysisof consumption (XYZ); and (iii) A multi-criteria perception of risk(AHP). Therefore, both qualitative and quantitative criteria wereintegrated in the proposed method (see Fig. 1).

2.1. ABC analysis

The ABC inventory control method is a helpful technique fordetermining which inventories should be counted more frequentlyand managed more closely than others (Ramanathan, 2006;Al Kattan and Bin Adi, 2008). The ABC classification method showswhere the efforts in handling the inventories can be best appliedand where the best opportunities can be found to reduce costs

360 M.G. Cedillo-Campos, H.O. Cedillo-Campos / Safety Science 79 (2015) 358–368

(Flores and Whybark, 1988; Hadi-Vencheh, 2010). The aim of thisselective system is to reduce time, effort and cost in controllinginventories (Al Kattan and Bin Adi, 2008; Jin-Xiao, 2011;Millstein et al., 2014). From a practitioner point of view, in manycases, the classification of SKU is only based on one criterion, gen-erally the relationship between cost and use (Lung, 2006;Cedillo-Campos and Sánchez, 2013). Thus, Flores and Whybark(1988) suggest making a two-step classification including multiplecriteria and specific rules to classify SKU. On the first step, from adisruption risk perspective, a direct classification of the SKU ismade, according to the most critical ones. On the second step, eachsupply policy is selected for each type of SKU, looking for theappropriate way to handle each article of the inventory. Thus,the items of the inventory are divided in three types according totheir value and importance. This division shows the articles thatrequire more control, and are classified as follows:

Type A: Articles that require a 100% strict control.Type B: Articles that require less effort and a lower administra-tive cost.Type C: Articles that only require a simple supervision.

In order to classify SKU according to their inventory turnover,the following steps are usually used:

1. Getting the average sales from the month sales of each SKU.2. Ranking the average obtained (monthly sales) from less to

more.3. Assigning a SKU code of the ordered sales, a sale rank for each

SKU and the average of ordered monthly sales.4. Adding up the total average of monthly sales.5. Getting the percentage of monthly sales by dividing the average

of monthly sales of each SKU code by the total of average sales.6. Defining the percentage of articles dividing one by the total

articles.7. Establishing the percentage of accumulated sales and the

percentage of accumulated articles.

2.2. XYZ analysis

According to Wildemann (2001) and Schönsleben (2007), theXYZ classification enables realizing the following step to the inven-tory analysis thanks to the ABC selective system, but focusing onthe variability for each subgroup. As such, it is possible to carryout a detailed analysis using the XYZ, taking into account theregularity of consumption for each SKU. The classification accord-ing to the variability is divided as follows:

Materials X: Articles with regular consumption. The variationsinside this category are small and the forecast related to thisclassification is usually accurate.Materials Y: Articles with an irregular consumption, hard toforecast and material with increasing and decreasingtendencies.Materials Z: Articles with a very irregular consumption.Variability is important and forecast is difficult.

The combination of the ABC and XYZ systems represents anABC–XYZ matrix (see Table 1). This is the first step to build themethod that combines both results, providing important informa-tion. With this information as a basis, a practitioner can defineaccurate measures to optimize inventories. For the variabilityanalysis, the analytic steps are organized as follows:

1. The SKU code and the classification from the ABC analysis arere-written in the same order; that is to say, they are classifiedaccording to the volume of sales (V). The data of every SKU bymonth is considered in a descending order.

2. A total volume of sales (V) is made for each month, which isobtained as follows:

Vj ð1Þ

3. An average (A) is obtained making sure it coincides with themonthly volume of sales (V) of the ABC classification:

A ¼Pn

j¼1Vj

Nð2Þ

4. The standard deviation (s) for each month is calculated asfollows:

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiXn

ðVj � VÞ2

N � 1

sð3Þ

5. The coefficient of variation (CV) is obtained by dividing thestandard deviation (s) by the average monthly sales (A):

CV ¼ s

A� 100 ð4Þ

Based on the combined ABC–XYZ matrix, the results of the twoanalyses can be combined and thereby, important information onmaterials and inventories can be obtained. Actually, it enablesdefining appropriate measures for optimizing inventory. Hence,this matrix allows choosing the appropriate method of materialmanagement for individual items (see Table1).

2.3. AHP risk analysis

As Torabi et al. (2012) argue, companies typically use the ABCanalysis based on just one criterion, i.e., the annual dollar usageto classify inventory items. However, most of the current modelsassume that all criteria are of quantitative type and hence cannothandle the qualitative ones (Badeaa et al., 2014). Thus, a moreeffective ABC inventory classification needs to integratemulti-criteria approaches in the case where there would be bothquantitative and qualitative criteria. At the same time, as Marchand Shapira (1987) suggest, risk can be understood as a ‘‘variationin the distribution of possible outcomes, their likelihoods and theirsubjective values’’. In that sense, risk is a speculative component(Pfohl et al., 2010). Consequently, it became essential to integrateAHP as a methodological approach capable to ponder quantitativeand qualitative criteria.

Largely used in many different areas, the AHP (AnalyticHierarchy Process) multi-criteria decision method developed byThomas Saaty in the 1980’s, counts with solid mathematical andpsychological foundations. In fact, Ho (2008) and Lollia et al.(2014) stand that AHP is a powerful solution tool for those prob-lems where considering quantitative values is not enough, but alsorequires qualitative factors and perceptions based on the experi-ence. It is the case of risk classification of SKU located in awarehouse.

Huan et al. (2004) as well as Rodrigues et al. (2014) argue thatthe AHP was developed to reflect the natural shape of people’smind and behavior. The AHP is designed to solve multi-criteriacomplex problems, and can be applied on many fields (Ho, 2008;Rex et al., 2014). This method is based on the decomposition of acomplex problem into a hierarchic structure of different levels,

Table 1The ABC–XYZ matrix (Schönsleben, 2007).

Continuousness ofdemand

Consumption value

A High B Medium C Low

X High High value + continuous demand Medium value + continuous demand Low value + continuous demandY Medium High value + regular or fluctuating demand Medium value + regular or fluctuating demand Low value + regular or fluctuating demandZ Low High value + discontinuous demand Medium value + discontinuous demand Low value + discontinuous demand

where the objectives, the criteria, the sub-criteria, and the alterna-tives can be found (Berumen and Llamazares, 2007; Rodrigueset al., 2014). It actually implies to identify resources, skills, and val-ues. It also tries to transfer perceived reality by the individual to ascale of reason, where priorities related to each element arereflected (Saaty, 1980; Fuh-Hwa and Lin-Hai, 2005; Lollia et al.,2014).

The AHP enables synthesizing efficiently and graphically theinformation regarding a problem, decomposing it, and analyzingit by parts, prioritizing and evaluating each element through bin-ary comparisons according to its importance. As a result, it is usefulto make decisions based on a mathematical support and to includethe risk perceptions of the actors involved in a participative pro-cess, enabling the AHP solutions to complement other mathemat-ical optimization methods. The AHP’s aim is to integrate differentmeasures into a single overall score to rank decision alternatives(Deng et al., 2014).

Even if authors as Hazelrigg (2012) argue that AHP can beinconsistent, other authors agree that it remains a reliable andwidely used method in multiple-attribute decision making(Mosadeghia et al., 2015; Oztaysi, 2014). In fact, results providedby different authors demonstrate that in the stage of the identifica-tion process, simplified methods can be sufficient. In this situation,selecting more sophisticated techniques will not necessarily gener-ate different outcomes (Rezaei et al., 2014; Srdjevic and Srdjevic,2014; Taylana et al., 2014).

Furthermore, since AHP is well known between decisionmakers in emerging markets as the Latin-American one, andunderstanding clearly a method is a critical factor of successwhen implementing it to real-world problems, this method wasselected.

2.3.1. Structure of the modelThe structure of w@reRISK method (see Fig. 2) describes the

level of security risk for the classification of SKU in a warehouse.The sub-criteria on two levels arises from the characteristics ofthe SKU and its use. The alternatives are high risk and low risk.

2.3.2. Judgments and evaluationsAt this stage, the alternatives are evaluated using pairwise com-

parisons for each of the risk criteria. Specifically, it develops aPairwise Comparison Matrix (PCM) of the alternatives for each ofthe criteria, establishing the importance, preference, or likelihoodby assigning a numerical value that measures the intensity of pref-erence and relative importance between the two alternatives con-sidered. This value is extracted from Table 2, making the verbalscale of comparison, the equivalent numerical scale as the scaleof Saaty (1980). In fact, the matrix of pairwise comparisons con-tains alternative comparisons or criteria.

The preferences were assigned performing a Nominal GroupTechnique (NGT), which allows to vote in an anonymous way,and providing equal opportunities to participate to all group mem-bers. However, as any technique, some disadvantages weredetected: (i) opinions may not converge in the assigning process

of preferences; and (ii) cross-fertilization of perceptions may beinhibited (which could be an advantage too).

The first level of criteria corresponds to the physical character-istics and use of the SKU. The second level states the risk directlyrelated to the physical characteristics of the SKU and the potentialillegal uses of it. Finally, the third level of criteria identifies thespecific criterion to be considered when locating SKU in a ware-house. The sub-objectives were chosen after brainstorming withexperts and an extensive literature review of scientific and practi-cal papers as well as security initiatives as C-TPAT, BASC (BusinessAlliance for Secure Commerce) or standards as ISO 31000. In theinterest of facilitating a clear understanding of the factors used,these are presented in Table 3.

If we consider a square matrix A with dimensions n � n (5), withrelative judgments on the criteria and if aij is the element (i, j) of A,for i = 1, 2, . . . n, and j = 1, 2, . . . n, it can therefore be affirmed that Ais a matrix of paired comparisons of n criteria, if aij is the measureof preference for the criterion from row i when it is compared tothe criterion from column j:

1 a12 � � � a1n

a21 1 . . . a2n

. ... . .

an1 an2 . . . 1

266664

377775 ð5Þ

When i = j, the value of aij will be equal to 1, as a criterion, it iscompared to itself. This is how (6) is fulfilled:

aij � aij ¼ 1 : A ¼

1 a12 � � � a1m

1 � � � a2m

. ... . .

� � � 1

2666664

3777775 ð6Þ

In matrix A, all elements are positive and satisfy the followingproperties:

i. Reciprocity: if A is a matrix of paired comparisons, it is ful-filled: aij = 1/aij, for all i, j = 1, 2, . . . n.

ii. Consistency: where aij = aik/aik for all i, j, k = 1, 2, . . . n.

Each cell in the matrix will have one of the values of Saaty scale,where comparisons located to the left of the diagonal of the value 1have a preference intensity located versus the right side of thediagonal.

2.3.3. Normalization of the matrixGiven the comparison matrix (5), the elements of each column

are added vertically, getting the values:

v1;v2; . . . ;vn ¼Xn

ai ð7Þ

Once the sum of each column has been made, each element ofthe matrix is divided between the sums obtained to get the stan-dard comparison matrix (8):

Security Risk Level Classification of SKU(w@reRisk)

Characteristics Use

Damage tothe product

Damage toStaff

PropertyDamage

Terrorism MaterialSmugglig

HumanSmuggling

Temperature

Humidity

Fragility

4 Criteria

High Risk Low Risk

7 Criteria 6 Criteria 8 Criteria 4 Criteria

Corrosive

Sharp-edged

Specializedhandling

Difficult tomaneuver

VoluminousValuable

Movable

Modifiable

Concealable

Marketable

Attractive

Nuclear

Bological

Chemical

Informatics

Radiological

5 Criteria

Unconventionalweapons

Tools which could beused as weapons

Chemicals

Pharmaceuticals

Blunt instruments

Illegal drugs

Sharp-edged

Conventionalweapons and/ortheir components

People

Plants and/orseeds

Animals and/orderivatives

OrgansLight

Fig. 2. Hierarchy of security risk of SKU.

Table 2Scale of measurement of preferences between two elements (Saaty, 1980).

Degree ofpreference(aij)

Verbal Judgment Description

1.0 Both elements are equallyimportant

Both elements equallycontribute to the property

3.0 Moderate importance of The experience and opinion

Anormalized ¼

a12v2� � � avn

1v2� � � a2n

. ... . .

an2v2� � � 1

26666664

37777775

one element over another favor one element over another5.0 Strong importance of one

element over anotherAn element is strongly favored

7.0 Very strong importance ofone element over another

An element is stronglydominant

9.0 Extreme importance of oneelement over another

An element is favored, at leastwith an order of magnitude ofdifference

2.0, 4.0, 6.0,8.0

Intermediate valuesbetween two adjacentjudgments

Used as values of consensusbetween two judgments

2.3.4. Development of the criteria priority vectorFor priorities from the standardized matrix, column vector con-

taining the averages of the rows of the standard matrix is com-puted as follows:

266666666666664

377777777777775

This row average represents the Priority Vector alternative tothe criterion considered:

266664

377775 ð10Þ

It is necessary to verify that the sum of the priority vector ele-ments is equal to 1, i.e. corresponding to 100%.

Table 3Description of selected factors.

Factor Description

� Characteristicsof SKU

Risk associated with specific physical SKU characteristics

� Use Risk arising from the use (from a maliciousnessperspective) of the SKU itself

� Damage to theSKU

Due to the physical or chemical characteristics of the SKUwith the risk to be damage or destroyed as a result of ahandling or storage process

� Damage tostaff

SKU characteristics that make it dangerous to the staffhandling it

� Propertydamage

Specific SKU characteristics or perception of them thatincrease the risk of a SKU to be stolen

� Terrorism Use of a SKU to produce terror as a mean of coercion� Material

smugglingSmuggling of goods is the furtive shipping of goods ormaterials that past a point where outlawed, such as aninternational border, in contravention of related laws orother conventions

� HumanSmuggling

Human Smuggling is the undercover shipping of beingspast a point where outlawed, such as an internationalborder, in contravention of related laws or otherconventions

� Valuable SKU having advantageous or admired characteristics orqualities

� Attractive SKU that can produce affection, satisfaction, or pleasure� Concealable Refers to SKU that are easy to hide, or hard to identify once

stolen� Moveable SKU that can be easily moved from one place to another� Modifiable SKU that can be used to transport illegal SKU, substances,

or others� Marketable SKU that can be quickly sold

2.3.5. Analysis of consistency of the opinionsInconsistency arises when some judgments of the pair wise

comparison matrix are inconsistent with others. Human judg-ments are often imperfect, so it is very difficult to have accuratemeasurements for wi. However, the AHP allows measuring thedegree of inconsistency at this stage; thus, that consistency hassimultaneously two properties:

a. Transitivity of preferences, which indicates that the judg-ments should respect traffic conditions caused by comparingmore than two elements, i.e., if w1 is better than w2, and w2

is better than w3, then it is expected that w1 is better thanw3.

b. The proportionality of preferences which indicates proportion-ality, i.e., if w1 is three times as good as w2, and w2 is twice asgood as w3, then w1 is expected to be six times as good asthan w3.

According to these properties, we can say that a matrix An�n isconsistent when pair wise comparisons are based on exact mea-surements, i.e. when the values w1 . . . wn are known and obtainedaij = wi/wj. If we take the paired comparisons between alternativesin the matrix A (5), we can say that the element a12 is the impor-tance between Alternative 1 and Alternative 2. By making an anal-ogy of values, and assuming that alternative 1 takes the value of w1

and 2 takes the value of w2, computed as follows:

a12 ¼w1

w2ð11Þ

If in the matrix each element aij is replaced by the ratio w1/w2,we will have the following matrix:

ðAÞ ¼

w1w2� � � w1

w2w2� � � w2

. ... . .

wnw2� � � wn

2666664

3777775 ð12Þ

If we consider the line i of the matrix of judgments: ai1, ai2, -. . . aij, . . . ain, in the ideal case, and we multiply the elements ofthe line w1, w2, . . . wn, we will have the following:

wi=w1 �w1 ¼ wi wi=w2 �w2 ¼ wi

wi=wj �wj ¼ wi wi=wn �wn ¼ wi

ð13Þ

Consistency Ratio (CR) (14) of the opinions used in the pair wisecomparison matrix can be determined with AHP, calculating theconsistency ratio as an index of consistency (CI) of (A) and the ran-dom consistency index (AI):

CR ¼ CIAI

ð14Þ

where the deviation of the consistency is represented by the consis-tency index (CI), which measures the dispersion of the judgments ofthe practitioner of matrix A and is obtained as follows:

CI ¼ kmax � nn� 1

ð15Þ

where the random consistency index (AI) or random index (RI) istaken from Table 4 with values already determined. It is consideredthat the consistency of the practitioner is acceptable whenCR < 0.10. For those cases where CR > 0.10, the opinions and judg-ments should be reconsidered. The steps for determining the ratioof consistency are:

1. For each line of the pair wise comparison matrix, determining aweighted sum based on the addition of SKU from each cell bythe corresponding priority of each alternative.

2. For each line, dividing the weighted sum by the priority of thecorresponding alternative. The result is kmax.

3. Calculating the consistency index (15) for each alternative.4. Rating RI, or the AI in Table 4.5. Determining the consistency ratio (CR) (14).6. If CR is lower than 0.10, go to step 6; otherwise, the opinions

expressed should be reconsidered.

2.3.6. Development of matrix prioritiesThe results obtained in step 4 for each criterion are summarized

in a Priority Matrix (MP) (16), listing the alternatives by the rowand column criteria. To obtain the priorities of the alternatives,the matrix that contains the priorities of the alternatives againstthe criteria at all levels has to be calculated.

p11 p12 � � � p1m

p21 p22 � � � p2m

. ... . .

pn1 pn2 � � � pnm

266664

377775 ð16Þ

2.3.7. Development of the global priority vectorComparison Criteria Matrix (17) in pairs is developed like those

made for alternatives in stages 2, 3, 4 and the review of step 5:

Table 4Random consistency index.

Number of elements of the matrix 1 2 3 4 5 6 7 8 9 10

Random consistency index (AI) 0.00 0.00 0.58 0.89 1.11 1.24 1.32 1.40 1.45 1.49

ð17Þ

The matrices obtained are multiplied by the matrices of the vec-tors of priorities for the sub-criteria (18) compared to ahigher-level approach:

p11 p12 � � � p1m

p21 p22 � � � p2m

. ... . .

pn1 pn2 � � � pnm

266664

377775

266664

377775 ¼

266664

377775 ð18Þ

The process is repeated until the end of all comparisons of themodel elements.

2.3.8. ResultsThe results based on priorities, judgments and assessments

made through the comparisons of the components of the hierarchi-cal model are shown in an alternatives ordering that are organizednumerically by their risk level. For this purpose, alternative risklevel will be high-risk and low-risk so the major alternative of50% determines the risk level for each SKU. In the event that bothalternatives are exactly 50%, the risk level will be medium.

2.4. An integrated approach

By integrating the different approaches mentioned above, notonly is it possible to know the inventory turnover and variabilityof the inventories but also to evaluate the risk level. As a result,since organizational efficiency and performance are enhancedwhen tools to reduce uncertainty take into account ‘‘context’’ and‘‘environmental realities’’ (Duncan, 1972), a flexible method wasdeveloped, which can correctly classify SKU in warehouses regard-ing security risk, inventory turnover and variability. The purposewas also to consolidate a technically solid analytical framework,

Table 5An ABC classification.

easy to use by practitioners through a spreadsheet in their taskto improve supply chain security. Even though reactive initiativesas C-TPAT are useful, proactive initiatives that develop solutionsfor a formalized approach to security management within the firmare needed (O’Connell, 2009; Hintsa et al., 2010).

3. Illustrative example

The variability of the different SKU, their physical and economicsimilarities and differences, as well as their sale behavior accordingto seasons and trends, only to mention some of them, are elementsto consider for an accurate classification of SKU. Not consideringthe inherent risk to the SKU (see Fig. 2) limits the effective opera-tion of a modern warehouse.

As an example, from a security risk perspective (based on a prop-erty damage concern) we performed the w@reRISK method withsupport of a well-known logistics company with facilities in differ-ent regions around the world. A group of eight SKU (informaticscomponents) was chosen to perform the ABC classification as wellas the XYZ variability analysis. However, looking for a less complexpresentation of the method, for the final step involving the multicri-teria risk analysis, only one SKU was analyzed. Since the same pro-cess was achieved to all of the SKU, the analysis based on one SKUappeared enough to prove the utility of the proposed method.

3.1. ABC classification of SKU

In this first stage, a set of localized SKU in a specific section ofthe company’s warehouse was selected for our analysis. The SKUsample consisted of 8 SKU of which 21% represented 75% of thecompany sales. Thus, the SKU were prioritized into three groupsbased on inventory usage (used Quantity � Price), determiningthe average inventory per group (see Table 5).

3.2. XYZ variability analysis

Focusing on the variability for each subgroup defined by theABC classification, it was possible to realize a complete study bymeans of XYZ approach. The value of the access fluctuations was

Table 6The XYZ classification.

determined using a fluctuation coefficient. This coefficient repre-sents the deviation of the access pattern in the present period fromthe one in the previous period. As Hoppe (2006) explains, ‘‘if thefluctuation coefficient increases, the forecast accuracy decreases’’.Thus, for this case, ‘‘X’’ materials have a fluctuation coefficient of<0.1; ‘‘Y’’ materials have a fluctuation between 0.1 and 0.25; andthe fluctuation of ‘‘Z’’ materials are >0.25. From this approach,three X SKU, three Y SKU, and two Z SKU were identified (seeTable 6).

3.3. AHP risk analysis

The AHP Risk analysis was developed for each of the eight SKUconsidering all sub-objectives presented in Fig. 2. However, toshow synthetically the advantages of the proposed method, thedeveloped steps were shown only to analyze the risk level of oneSKU. The SKU A1 is one of the ‘‘star’’ SKU of the company whichis perceived by practitioners with a risk level that needed to be bet-ter measured. It is a highly demanded SKU, but it is also highly fluc-tuating (see Table 6).

In order to improve agility in the method implementation, aquestionnaire regarding risk perceptions with an easy answeringformat was developed by the warehouse manager. Once collected,the perceptions were entered into a spreadsheet in MS Office Excel� for integrated analysis. For all the analysis, we obtained a relation-ship, CI

RI < 0:10 allowing us to guarantee the validity of the results.Although each SKU had different levels of total risk, it was possibleto assess the risk for each of the sub-objectives, which is actuallyvery valuable because identifying the type of higher risk mayimprove the way the SKU are stored and handled in a warehouse.

3.3.1. Comparison by pairs of the matrixTo assess the alternatives, the matrix of comparison by pairs of

alternatives for each criteria and sub-criteria is developed,

Table 7Paired comparison of the alternatives, ‘‘valuable’’ criterion of 3rd level.

High risk Low risk

High risk 1.00 5.00Low risk 0.20 1.00

Table 8Paired comparison of the criteria to assess the damage to the patrimony.

Concealable Moveable Valua

Concealable 1.00 1.00 7.00Moveable 1.00 1.00 5.00Valuable 0.14 0.20 1.00Attractive 0.20 0.20 5.00Marketable 1.00 1.00 3.00Modifiable 0.11 0.14 0.33

establishing the importance, preference, or probability. Based ona NGT technique, practitioners working in the warehouse assignedthe preference of the alternatives by granting a numerical value toeach comparison (see Table 3). In Table 7, the data indicates therisk perception of A1 SKU, and why it is considered as valuable.

The data of Table 8 indicate the importance of ith criterionwhen comparing it versus jth criterion of ‘‘property damage’’.

3.3.2. Standardizing the matrix of comparison by pairsAfter the matrices of comparison by pairs have been obtained,

the following step is to calculate the standard matrices. In orderto standardize the matrix, each entrance in column i of A is dividedby the sum of the entrances in column i for all the levels of criteria.In Table 9, the standardized matrix is calculated from Table 10.

3.3.3. Developing the priority vector for each criterionThe column vector that contains the averages of the rows of the

standard matrix is calculated (to see Table 9). This average for eachrow represents the priority vector of the alternative related to theconsidered criterion (see Table 10).

Then, one verifies that the sum of the elements of the priorityvector is equal to 1, that is to say, that it corresponds to 100%. AsFig. 3 shows, the priority for the evaluation of damage to the pat-rimony is the ‘‘concealable’’ criterion, followed by ‘‘moveable’’ oneand ‘‘marketable’’ one. Fig. 3 also shows that the smaller prioritycorresponds to the fact that the SKU is ‘‘modifiable’’.

3.3.4. Analyzing the consistency in the comparison by pairsThe inconsistency arises when some judgments of the matrix of

comparison by pairs contradict with each other. It is very compli-cated to have exact measures when the values that compare thecriteria are settled down. Indeed, the human judgments are usuallyimperfect. For example, the practitioners can decide that A has ahigher value than B, and that B has a superior value than the oneof C, but the practitioners may state that C is higher than A. It isthen when an inconsistency occurs, since being based on the pre-vious comparisons, AHP expects A to be higher than C. Thus, AHPincorporates the inconsistencies to the model.

The consistency has two simultaneous properties: the transitiv-ity of the preferences and the proportionality of the preferences.The first points out that the emitted judgments must respect theoriginated conditions of transitivity when comparing more than

ble Attractive Marketable Modifiable

5.00 1.00 9.005.00 1.00 7.000.20 0.33 3.001.00 0.33 5.003.00 1.00 9.000.20 0.11 1.00

Table 9Standard Matrix of the criteria to assess the damage to the patrimony.

Concealable Moveable Valuable Attractive Marketable Modifiable

Concealable 0.29 0.28 0.33 0.35 0.26 0.26Moveable 0.29 0.28 0.23 0.35 0.26 0.21Valuable 0.04 0.06 0.05 0.01 0.09 0.09Attractive 0.06 0.06 0.23 0.07 0.09 0.15Marketable 0.29 0.28 0.14 0.21 0.26 0.26Modifiable 0.03 0.04 0.02 0.01 0.03 0.03

Table 10Priority vector.

Risk of damage to the patrimony

Concealable 0.30Moveable 0.27Valuable 0.06Attractive 0.11Marketable 0.24Modifiable 0.03

Total 1.00

Modifiable

Marketable

Attractive

ValuableMoveable

Concealable

0.00 0.10 0.20 0.30 0.40

Fig. 3. Priority of the criteria ‘‘property damage’’.

Table 12Priority matrix of the criteria of level 1 to evaluate the risk level of SKU A1.

Characteristics Use of the SKU Results

High risk 0.80 0.40 0.75Low risk 0.20 0.60 0.25

two elements and the second one implies proportionality besidestransitivity. The index of consistency (IC) for the twin matrix, inthis case is of 0.05706, and the random index (RI) is of 1.24. Therelation of consistency in this case is of 0.04601, minor to 0.10;therefore, the consistency degree is satisfactory.

3.3.5. Developing the matrix of prioritiesThe results obtained in stage 4 for each criterion are summa-

rized in a matrix of priorities. The matrix of priorities for ‘‘risk ofdamage to the patrimony’’ for level 3 of criteria appears inTable 11.

3.3.6. Vectoring global priorityThe results of Table 11 were obtained by multiplying the prior-

ity matrix for ‘‘risk of damage to the patrimony’’ by the priority vec-tor ‘‘risk of damage to the patrimony’’ (see Table 10). The matrix ofpriority of levels two and one are calculated likewise. When all thecomparisons of the elements of the model are finished, the resultwill be the global risk. In Table 12, the result of high risk is equiv-alent to 75%, and the result of low risk is equal to 25%. It is con-cluded that SKU A1 has a high-risk level.

Consequently, resources (economic and human) should be allo-cated to design a product-oriented risk mitigation plan. As shownin Fig. 3, the analyzed SKU A1 requires a risk mitigation plan more

Table 11Priority of the criteria of level 3 to evaluate ‘‘risk of damage to the patrimony’’ matrix.

Concealable Moveable Valuable

High risk 0.9 0.9 0.83Low risk 0.1 0.1 0.17

focused on the ‘‘property damage’’, mainly taking into account the‘‘concealable’’ criterion of the SKU, followed by ‘‘moveable’’ one,and then ‘‘marketable’’ one.

4. Concluding remarks

Based on a comprehensive literature and field practices review,we concluded that, from a balanced approach including at thesame time a solid theoretical background and practical processesof implementation, most of current methods to manage ware-houses still do not take into account the risk level classificationof Stock Keeping Units (SKU). In that sense, we count three contri-butions of the paper to the practice as well as the body ofknowledge.

First, from a methodological approach, since modern risk anal-ysis involves factual data, but also subjective values, our contribu-tion provided an integrated analytical framework, which involvedthe use of the ABC classification of SKU, the XYZ variability systemanalysis, and the AHP to ponder quantitative and qualitative val-ues. These three methods, from an integrated approach, not onlyenable taking into account elements such as the relationshipbetween cost and use or the variability to classify SKU, but alsoconsidering the risk level of each SKU when locating it in awarehouse.

Second, in the last few years, the new means to mitigate risksgenerated on the supply chains by terrorist activities have receivedmuch attention. Still, in the emerging markets, the security supplychain policies are more driven by crime than by terrorism. In thatsense, there was a lack of practical and academic propositions inte-grating both approaches. Consequently, the method here proposedcontributes to a better design process when designing global sup-ply chains from a multidimensional perspective of risk.

Finally, thanks to its global approach, the method can not onlybe used to classify SKU, but also any kind of physical items (pallets,containers, etc.) that must be stored or located in a physical space(terminals, logistics platforms, etc.). The generalist approach of themethod provides an important flexibility that can be adapted tothe specific needs of the user. Furthermore, due to its logical

Attractive Marketable Modifiable Results

0.87 0.9 0.17 0.870.13 0.1 0.83 0.13

approach, it is susceptible to become an automated informaticstool to easily reduce warehouses’ security breaches.

One of the limitations of this work would be that prior to runthe proposed method, the decision makers must accept the impor-tance of the risk related to the logistics operations. Not all organi-zations perceive the risk and thus, do not count with a clear risklevel classification for each product. As such, the benefit of theproposed method can be reduced. On the other hand, from a prac-tical point of view, even though using the AHP as a multi-criteriamethod happens to be relevant since decision makers in LatinAmerica well know it, other approaches such as Fuzzy-AHP shouldbe considered to improve the tool effectiveness (Isaai et al., 2011;Cho et al., 2012; Podgórski, 2014). Indeed, the method hereproposed complements the warehouse management policies theorganizations already implemented.

4.1. Future research

A security breach not only causes financial consequences andnegative impacts on trademark prestige for a company that suffersit. From a more comprehensive perspective, it also drops economicconfidence for doing business with members that are part of thesame supply chain. Thus, the perception of potential risk propaga-tion is not currently well analyzed and will be part of a futureresearch. Also, the impact on the quality of the supply chain prod-ucts/processes of the proposed method will be made. Furthermore,a mobile informatics tool based on internet is being worked on,which would enable an easy implementation of the proposedmethod, with a user-friendly interface. By using a mobile device,survey information about risk perception is one of the actions thatwill make this method faster. Form a technical point of view, tostrengthen our approach; work is in process to deal with theintegration of the Fuzzy Logic approach in the evaluation ofperceptions.

Funding

The authors acknowledge all the support provided by theNational Council of Science and Technology of Mexico(CONACYT) through the research program ‘‘LaboratoriosNacionales’’ (National Laboratories) as well as by the MexicanLogistics and Supply Chain Association (AML), and the MexicanInstitute of Transportation (IMT).

Acknowledgements

The authors thank Flora Hammer for several useful commentsthat improved this paper. A special thanks to Ernesto Donnadieu,Operations Director Mexico at Ryder Integrated Logistics for hissignificant comments that improved our research approach.However, all the imprecisions that could remain in the documentare only responsibility of the authors.

References

Al Kattan, I., Bin Adi, A., 2008. Multi-criteria decision making on total inventory costand technical readiness. Int. J. Interact. Des. Manuf. 2 (3), 137–150.

Badeaa, A., Prosteana, G., Goncalvesb, G., Allaoui, H., 2014. Assessing risk factors incollaborative supply chain with the analytic hierarchy process (AHP). Proc. –Soc. Behav. Sci. 124 (20), 114–123.

Bakshi, N., Gans, N., 2010. Securing the containerized supply chain: analysis ofgovernment incentives for private investment. Manage. Sci. 56 (2), 219–233.

Bakir, N., Pakdaman, D., 2006. An analysis of truck cargo security at Southwesternborders and potential impacts of cross-Border trucking services of Mexicancarriers on terrorism risk exposure. Center for risk and economic analysis ofterrorism events (CREATE). University of Southern California, Los Angeles.

Berumen, S., Llamazares, F., 2007. La utilidad de los métodos de decisiónmulticriterio en un entorno de competitividad creciente. Cuadernos deAdministración 20 (34), 65–87.

Bueno, A., Cedillo-Campos, M., 2014. Dynamic impact on global supply chainsperformance of disruptions propagation produced by terrorist acts. Transport.Res. Part E: Logist. Transport. 61, 1–12.

Buesa, M., Valiño, A., Heijs, J., Baumert, T., González, J., 2007. The impact of theterrorist attacks on the security market. In: A Comparative Study of theFinancial Impact of the Islamist Attacks on New York, Madrid and London.Información Comercial Española, No. 834.

Cedillo-Campos, M., 2012. Designing supply chains under security threat: the firstnational exploratory survey in Mexico. In Spanish. Technical Report to theSecretariat of Commerce and National Council of Science and Technology.México. <http://www.elogistica.economia.gob.mx/swb/work/models/elogistica/Resource/3/1/images/201>.

Cedillo-Campos, M., Perez, A., 2010. Hybrid supply chains in emerging markets: thecase of the Mexican auto industry. S. Afr. J. Ind. Eng. 21 (1), 193–206.

Cedillo-Campos, M., Sánchez, C., 2013. Dynamic self-assessment of supply chainsperformance: an emerging market approach. J. Appl. Res. Technol. 11 (3),338–347.

Cedillo-Campos, M., Sanchez-Ramirez, C., Vadali, S., Villa, J., Menezes, M., 2014.Supply chain dynamics and the ‘‘cross-border effect’’: the U.S.–Mexican border’scase. Comput. Ind. Eng. 72, 261–273.

Cho, D., Lee, Y., Ahn, S., Hwang, M., 2012. A framework for measuring theperformance of service supply chain management. Comput. Ind. Eng. 62,801–818.

Chopra, S., Meindl, P., 2013. Supply Chain Management: Strategy, Planning, andOperations, fifth ed. Prentice Hall.

Deng, X., Hu, Y., Deng, Y., 2014. Supplier selection using AHP methodology extendedby D numbers. Expert Syst. Appl. 41 (1), 156–167.

Donner, M., Kruk, C., 2009. Supply Chain Security Guide. World Bank, Washington,D.C..

Duncan, R., 1972. Characteristics of organizational environments and perceivedenvironmental uncertainty. Adm. Sci. Q. 1 (3), 313–327.

Elmoselhy, S., 2014. Hybrid lean–agile manufacturing system technical facet, inautomotive sector. J. Manuf. Syst. 32 (4), 598–619.

Flores, B., Whybark, D., 1988. Implementing multiple criteria ABC analysis.Engineering Cost and Production Economics. Publishers B.V., Amsterdam, pp.191–195.

Fuh-Hwa, F., Lin-Hai, H., 2005. The voting analytic hierarchy process method forselecting supplier. Int. J. Prod. Econ. 3, 308–317.

GAO – United States Government Accountability Office, 2008. Supply ChainSecurity: Challenges to Scanning 100 Percent of U.S.-Bound Cargo Containers.Testimony before the Subcommittee on Surface Transportation and MerchantMarine Infrastructure, Safety, and Security, Committee on Commerce, Science,and Transportation, U.S. Senate. GAO-08-533T, Washington, D.C.

GAO – United States Government Accountability Office, 2010. Supply ChainSecurity: DHS Should Test and Evaluate Container Security TechnologiesConsistent with All Identified Operational Scenarios to Ensure theTechnologies Will Function as Intended. Report to the Chairman, Committeeon Homeland Security, House of Representatives. GAO-10-887, Washington,D.C.

Hadi-Vencheh, A., 2010. An improvement to multiple criteria ABC inventoryclassification. Eur. J. Oper. Res. 201 (3), 962–965.

Hazelrigg, G., 2012. Fundamentals of Decision Making for Engineering Design andSystems Engineering, 527pp. ISBN 0984997601.

Hintsa, J., Ahokas, J., Männistö, T., Sahlstedt, J., 2010. CEN Supply Chain Security(SCS) Feasibility Study. CEN/TC 379 Supply Chain Security, Final Report,Economic Commission for Europe, United Nations, Genève.

Ho, W., 2008. Integrated analytic hierarchy process and its applications – aliterature review. Eur. J. Oper. Res. 186 (1), 211–228.

Holler, R., Schanck, J., 2010. Customs Modernization Handbook. AuthorizedEconomic Operator Programs. USAID Worldwide Support for Trade CapacityBuilding project.

Hoppe, M., 2006. Inventory Optimization with SAP�. SAP Press.Huan, S., Sheoran, S., Wang, G., 2004. A review and analysis of supply chain

operations reference (SCORE) model. Supply Chain Manage.: Int. J. 9 (1), 23–29.Isaai, M.T., Kanani, A., Tootoonchi, M., Afzali, H.R., 2011. Intelligent timetable

evaluation using fuzzy AHP. Expert Syst. Appl. 38 (4), 3718–3723.Jin-Xiao, C., 2011. Peer-estimation for multiple criteria ABC inventory classification.

Comput. Oper. Res. 38 (12), 1784–1791.Johansson, L., 2008. Supply Chain Security Initiatives: A Trade Facilitation

Perspective. Kommerskollegium, The National Board of Trade, Sweden.Jones, D., Tabor, J., Woodard, K., 2009. Combating supply chain crime: securing your

warehouse. In: CSCMP Annual Global Conference, Chicago, Illinois, USA, pp.20–23.

Kern, D., Moser, R., Hartmann, E., Moder, M., 2012. Supply risk management: modeldevelopment and empirical analysis. Int. J. Phys. Distrib. Logist. Manage. 42 (1),60–82.

Lee, H., Whang, S., 2005. Higher supply chain security with lower cost: lessons fromtotal quality management. Int. J. Prod. Econ. 96, 289–300.

Lollia, F., Ishizaka, A., Gamberini, R., 2014. New AHP-based approaches for multi-criteria inventory classification. Int. J. Prod. Econ. 156, 62–74.

Lung, W., 2006. A simple classifier for multiple criteria ABC analysis. Eur. J. Oper.Res. 177, 344–353.

March, J., Shapira, Z., 1987. Managerial perspectives on risk and risk taking. Manage.Sci. 33 (11), 1404–1418.

Markert, F., 1998. Assessment and mitigation of the consequences of fires inchemical warehouses. Saf. Sci. 30 (1–2), 33–44.

Millstein, M., Yang, L., Li, H., 2014. Optimizing ABC inventory grouping decisions.Int. J. Prod. Econ. 148, 71–80.

Mosadeghia, R., Warnkenb, J., Tomlinsona, R., Mirfenderesk, H., 2015. Comparison ofFuzzy-AHP and AHP in a spatial multi-criteria decision making model for urbanland-use planning. Comput. Environ. Urban Syst. 49 (2015), 54–65.

NBC, 2014. Huge haul of electronics stolen from fremont warehouses. <http://www.nbcbayarea.com/news/local/Fremont-Huge-Haul-Of-Electronics-Stolen-From-Warehouses-278427711.html>.

O’Connell, J., 2009. C-TPAT: major challenges. J. Transport. Secur. 2 (4), 137–147.Ochoa, C., 2010. Managing risk in the global supply chain. U.S. Customs and Border

Protection. In: International Conference on Supply Chain Security, Monterrey,Mexico.

Oztaysi, B., 2014. A decision model for information technology selection using AHPintegrated TOPSIS-Grey: the case of content management systems. Knowl.-Based Syst. 70, 44–54.

Pfohl, H.-C., Kohler, H., Thomas, D., 2010. State of the art in supply chain riskmanagement research: empirical and conceptual findings and a roadmap forthe implementation in practice. Logist. Res. 2 (1), 33–44.

Podgórski, D., 2014. Measuring operational performance of OSH managementsystem – a demonstration of AHP-based selection of leading key performanceindicators. Saf. Sci. 73, 146–166.

Ramanathan, R., 2006. ABC inventory classification with multiple-criteria usingweighted linear optimization. Comput. Oper. Res. 33, 695–700.

Rex, E., Wu, T., Shunk, D., 2014. A stochastic AHP decision making methodology forimprecise preferences. Inf. Sci. 270 (20), 192–203.

Rezaei, J., Fahim, P., Tavasszy, L., 2014. Supplier selection in the airline retailindustry using a funnel methodology: conjunctive screening method and fuzzyAHP. Expert Syst. Appl. 41 (18, 15), 8165–8179.

Rodrigues, F., Osiro, L., Ribeiro, L., 2014. A comparison between Fuzzy AHPand Fuzzy TOPSIS methods to supplier selection. Appl. Soft Comput. 21,194–209.

Saaty, T., 1980. The Analytical Hierachical Process. J. Wiley, New York.

Schlegel, G., Trent, R., 2012. Risk Management: Welcome to the New. LogisticsManagement, pp. 42–45. <http://logisticsmgmt.com/images/site/LM1202_RiskManagement.pdf> [Consulted 03-18-2012].

Schönsleben, P., 2007. Integral Logistics Management, third ed. AuerbachPublications.

Serdarasan, S., 2013. A review of supply chain complexity drivers. Comput. Ind. Eng.66 (3), 533–540.

Sodhi, M., Son, B., 2012. Researchers’ perspectives on supply chain riskmanagement. Prod. Oper. Manage. 1 (21), 1–13.

Srdjevic, B., Srdjevic, Z., 2014. Synthesis of individual best local priority vectors inAHP-group decision making. Appl. Soft. Comput. 13 (4), 2045–2056.

Taylana, O., Bafailb, A., Abdulaala, R., Kabli, M., 2014. Construction projects selectionand risk assessment by fuzzy AHP and fuzzy TOPSIS methodologies. Appl. SoftComput. 17, 105–116.

The White House, 2012. National Strategy for Global Supply Chain Security.Washington, U.S.A.

Thun, J.H., Hoening, D., 2011. An empirical analysis of supply chain risk managementin the German automotive industry. Int. J. Prod. Econ. 131, 242–249.

Torabi, S., Hatefi, S., Saleck Pay, B., 2012. ABC inventory classification in the presenceof both quantitative and qualitative criteria. Comput. Ind. Eng. 63 (2), 530–537.

Torabi, S., Soufi, H., Sahebjamnia, N., 2014. A new framework for business impactanalysis in business continuity management (with a case study). Saf. Sci. 68,309–323.

U.S. Overseas Security Advisory Council, 2014. Overall Crime and Safety Situation.<https://www.osac.gov/pages/ContentReportDetails.aspx?cid=15151>.

Vilko, J., Hallikas, J., 2011. Risk assessment in multimodal supply chains. Int. J. Prod.Econ. 140 (2), 586–595.

Wildemann, H., 2001. Das Just-in-Time Konzept – Produktion und Zulieferung aufAbruf [the just-in-time concept – production and delivery on demand], fifth ed.TCW, Munich.

Xue, L., Villalobos, R., 2012. A multi-objective optimization primary planning modelfor POE (Port-of-Entry) inspection. J. Transport. Secur. 5 (3), 217–237.

w@reRISK method: Security risk level classification of stock keeping units in a warehouse

Documents

Transcript of w@reRISK method: Security risk level classification of stock keeping units in a warehouse

Utility Poultry-Keeping - Forgotten Books

SAP Extended Warehouse Management 9.1 Release ...

FINAL NIT BHOPATPUR - Central Railside Warehouse ...

THE INTEGRATED METHOD OF WAREHOUSE LAYOUT ...

Warehouse Billing - Wiki

Tire warehouse plan advances - DigiFind-It

Champions Complete - Warehouse 23

Keeping it Vague - WUR eDepot

Warehouse Management

Roller bearing units

Implementing Warehouse Management Systems in Logistics

Data Warehouse Optimization with Hadoop - Informatica

Keepingbusinessinmotion, keeping lives in motion...

Warehouse Fulfillment and Distribution Center - Orderhive

DATA WAREHOUSE PART L: DATA CATALOG

chapter 4 - keeping clean

Warehouse 2020 Survey Analysis - Cirrus Logistics

Warehouse and Inventory Management Standard Operating ...

Gear units & Motors

ARENA-Series Passive Speakers - Warehouse Sound