Number18,2006 113 DECISIONSUPPORTSYSTEMS:ACASESTUDY INVESTELDURABLEGOODSMARKETING LutfuSagbansua UniversityofMississippi MIS/POMDepartment lutsua@gmail.com ABSTRACT Supply chain management d eals with the efficient coordination of enterprises along a value chain to providegoodsandservicesto endusers.Thesuccessinmanagingasupplychainheavilydependsonthe effectiveusageoftechnology.Decisionsupportsystems(DSS)playsucharole.ADSSassistsandsupportsthe humandecisionmakerinthedecisionmakingprocess.ImplementationofsuchaDSStoolbyVestelDurable GoodsMarketingintheirdistributionresourceplanningprocessisanalyzedandpresentedinthisstudy. Keywords:DSS,SupplyChain,InformationTechnology,Distribution TƏDARÜKİDARƏETMƏŞƏBƏKƏSİVƏVESTEL ELEKTRİKMALLARININSATIŞINDATƏTBİQİ XÜLASƏ Tədarükİdarəetmə Şəbəkəsi son istifadəçilərə xidmət etmək və malları tədarük etmək üçün səmərəli kordinasiyalimüəssisələril əbirgəəlaqəqurur.Ağırbirtədarükzincirinin idərəolunmasındamüvəffəqiyyət qazanmaq,effektlitexnologiyanınistifadəsindənirəligəlir.Qərardəstəksistemidəeləbucürbirroloynayır. Bir Qərar Dəstək Sistemi, qərar vermə mərhələsində insanın qə rar qəbul etməsinə yardim edir və onu dəstəkləyir.Vesteltərəfindənmalbazarıvəonlarınpaylanmaresursplanlarıbucürbirqərarverməsistem alətinintərəfindəntəhlilolunuraqbutətqiqatobyektindətəqdimolunur. Açar sözlər:Texnikitəchizat,informasiyatexnologiyaları,çatdırılma INTRODUCTION Many of the advances in the control and management of supply chains are driven by advancing computer technology. Supply chainmanagementproblemsarenotsorigid and well defined that they can be delegated entirelytocomputers.Instead,inalmostevery case, the flexibility, intuition, and wisdom that is a unique characteristic of humans is essential to manage the systems effectively. However, there are many aspects of these systems that can only be analyzed and understood effectively with the aid of a computer.Itisexactlythistypeofassistance whichdecision‐supportsystemsaredesigned to provide. As the name implies, these systems do not make decision, instead, they assistandsupportthehumandecisionmaker inhisorherdecision‐makingprocess. Decision‐support systems range from spreadsheets, in which users perform their own analysis, to expert systems, which attempt to incorporate the knowledge of expertsinvarious fieldsandsuggestpossible alternatives. The appropriate DSS for a particularsituationdependsonthenatureof the problem, the planning horizon, and the type of decisions that need to be made. In LutfuSagbansua JournalofQafqazUniversity 114 addition, there is frequently a trade‐off between generic tools that are not problem‐ specificandallowanalysisofmanydifferent kinds of data, and often more expensive systems that are tailored to a specific application. Within the various disciplines that make up supply chain ma nagement, DSSs are used to address various problems, from strategic problems such as logistic network design to tactical problems such as the assignment of products to warehouses and manufacturing facilities, all the way through to day‐to‐day operation problems like production scheduling, delivery mode selection, and vehicle routing. The inherent sizeandcomplexityof manyofthesesystems make DSSs essential for effective decision making.DSSinsupplychainmanagementare often called Advanced Planning and Scheduling systems. These systems typically coverthefollowingareas:Demandplanning, supply planning, manufacturing planning andscheduling. Typically, decision‐support‐systems use the quantifiableinformation available to illustrate various possible solutions, and allo w the decisionmakertodecidewhichoneisthemost appropriate, based on other, possibly non ‐ quantifiable factors. Often, DSSs allow the decisionmakertoanalyzetheconsequencesof decision, depending on different possible scenarios. This kind of what ‐if analysis can helpavoidproblemsbefore theyoccur. Manydecision‐supportsystemsusemathema‐ tical tools to assist in the decision ‐making process. These tools, often from the mathe‐ maticaldisciplineofoperationsresearch,were firstdevelopedtoassistthearmedforceswith the enormous logistical challenges of World War II. Since th en, improvements in these techniques, as well as ever‐increasingcompu‐ terpower,havehelpedtoimprovethesetools andmakethemmoreaccessibletoothers. The to ols of art ificial intelligence are also employed in the design of decision‐support systems. Intelligent agents use AI to assist in decision making, especially in real‐time decision, suchasdetermininghowtosupplya customer in the shorte st possible time or to quote a delivery lead time as the cust o mer waitsonthephone.FollowingFox,Chionglo, and Barbuceanu, we define an agent as a softwareprocesswhosegoalistocommunicate and interact with other agents, so that decisionsaffectingtheentiresupplychaincan bemadeonagloballevel. SUPPLYCHAINDECISIONSUPPORT SYSTEMS Supply chain management encompasses a larger variety of decision. A list of such decisionsisprovidedbelow: ‐ DemandPlanning ‐ Logisticsnetworkdesign ‐ Inventorydeployment ‐ Salesandmarketingregionassignment ‐ Distributionresourceplanning ‐ Materialrequirementsplanning ‐ Inventorymanagement ‐ Productionlocationassignment/facility deployment ‐ Fleetplanning ‐ Leadtimequotation ‐ Productionscheduling ‐ Workforcescheduling SELECTINGASUPPLYCHAINDSS For each of the supply chain problems and issueslistedabove,decisionsupportsystems are available in many configurations, platforms, and price ranges. DSS platforms have evolved in the last 15 years from relatively inflexible mainframe systems, to isolated PC tools, to client/ server processes; lately, there is a new breed of high‐ performance and extensible enterprise decision‐support applications. These systems come in a wide range of pricing from PC systems costing several thousand dollars to company‐wide installations costing a few milliondollars. DecisionSupportSystems:ACaseStudyinVestelDurableGoodsMarketing Number18,2006 115 When evaluating a particular DSS, the followingissuesneedtobeconsidered: ‐ Thescopeoftheproblemaddressedbythe decision maker, including the planning horizon. ‐ Thedatarequiredbythedecision‐support system ‐ Analysisrequirements,includingaccuracy of the model, ability to quantify perfor‐ mance measures, desired analytical tools‐ thatis,optimization,heuristics,simulation, financial calculation requirements, and computationalspeedneeded. ‐ The system’s ability to generate a variety ofsolutionssothattheusercanselectthe most appropriate one, typically based on issuesthatcannotbequantified. ‐ The presentation requirements, including issues such as user‐friendliness, graphic interface, geographic abilities, tables, reports,andsoon. ‐ Compatibility and integration with existingsystems. ‐ Hardware and software system require‐ ments, including platform requirements, flexibility to changes, user interfaces, and technicalsupportavailable. ‐ Theoverallprice,includin gthebasicmodel, customization,andlong‐termupgrades. ‐ Finally,considercomplementarysystems. LITERATUREREVIEW Asupplychaincanbedefinedasanetworkof autonomous or semiautonomous business entities collectively responsible for procure‐ ment,manufacturinganddistributionactivities associated with one or more families of relatedproducts.Differententitiesinasupply chain operate subject to different sets of constraints and objectives. However, these entities are highly interdependent when it comes to improving performance of the supplychainintermsofobjectivessuchason‐ time delivery, quality assurance and cost minimization. As a result, performance of any entity in a supplychaindependsontheperformanceof others, and their willingness and ability to coordinateactivitieswithinthesupplychain. A global economy and increase in customer expectations regarding cost and service have influencedmanufacturerstostrivetoimprove processes within their supply chains, often referred to as supply chain re‐engineering (Swaminathan,1996). Supply chain re‐engineering efforts have po‐ tential to impact the performance of supply chains.Oftentheyareundertakenwithonlya probabilistic view of the future, and it is essential to perform a detailed risk analysis before adopting a new process. In addition, many times these re‐engineering efforts are made under politically ad emotiona lly charged circumstances. As a result, decision support tools that can analyze various alternativescanbeveryuseful inimpartially quantifying gains and helping the organizationmakethe rightdecision(Feigin, An,Connors,andCrawford1996). The goals of supply chain management are design, operation and maintenance of integrated value chains to satisfy consumer needs in the most efficient way by simultaneously maximizing customer service (Christopher,1998;Hewitt,1994;Ross,1998). Today, SCM is accepted as a concept integrating inter‐organizational business processesandcomprises other conceptssuch as Efficient Consumer Response, Quick Response, Continuous Replenishment and Customer Relationship Management (Bechtel and Jayaram, 1997). The design of supply chains requires the specification of business processes and supply chain wide planning routinesasspecialtaskofthedevelopmentof information systems as the backbone of any supply chain integration. Information technologyiswidelyperceivedastheenabler of supply chain integration (Bechtel and Jayaram, 1997; Hewitt, 1994). Enterprises participating as partners in a supply chain LutfuSagbansua JournalofQafqazUniversity 116 havetoprovidetheiractivitiesina waythat maximizes the supplychainefficiency.Thus, they have to concentrate on their core competencies(Christopher,1998). TheneedforDSScomesfromagapthatexists in the typical organization’s information resource management scheme. This gap is a clear indicator that classical data procession has not met the growing needs of modern businessconcerns.Forexample,today’schief executive is faced with an extensive list of fast‐developingproblems: - Thereisalargesetofincreasinglycomplex and comprehensive government agencies andregulationsimpactingonabusiness. - The economic climate has increased financialpressureonbusiness. - Many companies are now dealing in the world marketplace. With the improved capabilities of the transportation and communications industries, the business world has become smaller and more intensecompetitionhasresulted. Thesearesomeofthecurrentchallengesthat needtobeaddressed bybusiness. SUPPLYCHAINMANAGEMENTAT VESTEL Vestel Electronics A.S. is the largest electronics manufacturer in Turkey. Its core productTVswereaccountingfor70%oftotal salesin2000andmonitorsrepresented5%.In 2001,VestelElectronicsproducedatotalof4.6 million televisions, making up to 65% of the country’s total TV production. In 2002, TV productionincreasedto6.4million. While being a leading brand in the Turkish televisionmarketwith30%marketshareasof year2002VestelElectronicsisalsothelargest domestic brand exporter with 65% share. Being the largest full‐range television ODM (Original Design and Manufacturing) in Europe,VestelElectronicshadamarketshare of17%inOEMsales. VESTELDISTRIBUTIONNETWORK Most of the production occurs in a plant in Manisa. Imported goods are also received there. Until 1999, the company had four warehouses,servingthedealersandoutletsin different regions of the country. Distribution isperformedbyHorozLogistics.Withtheflat price per item pricing scheme given by the third‐party‐logistics (3PL) company, it was clear that there was no need to keep four warehouses. This led to an initiative of warehouse consolidation, whereby the distribution network took its current form with two warehouses. Other than the reductionindurablegoodsmarketcausedby thefinancialcrisisinTurkeyin2001,Vestel’s productionhasincreasedcontinuouslyasitis statedinthefollowingtable. Table1.ThenumberofUnitsShipped:Annualyand Monthly 2000 2001 2002 2003 Annual 900,000 518,867 592,652 1,007,701 Monthly 75,000 43,239 49,387 83,975 • In2001,duetothefinancialcrisisinTurkey,the durablegoodsmarketreducedby48% ANEWPLANNINGSYSTEM:MANUGISTICS TRANSPORTATIONMANAGEMENT Given the object ive of a better measurable system, Vestel decided to implement Manugistics’NetworkTransportManagement (MTM) module as the nex t improvement effortsforthedistributionsystemin2000.This package was chosen based on service options madeavailableinTurkeyby the various SCP providers and subsequent to anegotiation on price.VestelDurableGoodsMarketingwasthe firstcompanyinTurkeytoimplementsucha transportationplanningsystem,andremained theonlycompanyin2003. The distribution planning program is run daily to schedule deliveries to Vestel’s customers. The planning process is a part of theorderfulfillmentprocess: DecisionSupportSystems:ACaseStudyinVestelDurableGoodsMarketing Number18,2006 117 OrderEntry OrderAuthorization DistributionPlanning StockMovement Billing Distribution MTMCAPABILITIES MTMisatransportationoptimizationsoftware program, which provides the optimal route and truck planning for daily‐prepared deliveries. The inputs to the system are location of Vestel’s warehouses, transfer stations, and its customers; customer orders, transportation modes, and associated costs. The optimization program uses these inputs and finds a solution within the constraints imposedbythemanagementtominim izethe totaltransportationcosts.Therouteandtruck planningismadeaccordingtotheinputsand theconstraints. There are 3 different location types in MTM: warehouse, transfer station, customer. All the locationshavezipcodesgenera tedspecifically for MTM. These codes are different for each province.Somebigprovincesaredividedinto two or more regions. Th e distances between eachtwozipcodesareputinanetworktable. The distance between two points location in thesamezipcodeissettobe3km. VestelDurableGoods MarketingInc.hastwo warehouses, one in Manisa and the other in Istanbul. There are 9 regions throughout Turkey and the total number of transfer station in these regions is 19. The logistics companyownsandoperatesthesestations. The volume information for each product is providedasaninput intothesystem. Three different size trucks can be used for transportation in addition to a direct cargo alternative.Thecostsofusingeachalternative are set in the system. 10‐wheel or 8‐wheel trucks are used for the transportation to transfer station from the warehouses. Small trucks then make the deliveries from the transfer stations to the customers. There is also a direct cargo alternative from the warehouse in Manisa. Dealers with high volume demand can have direct deliveries with large trucks. MTM selects the direct cargo option based on transportation costs. Trucks utilizations constitute an important criterionfor decidingondeliverymode. Themanagementuses twopoliciesrelatedto efficiency and customer service. The first policy is related to truck utilization.A truck hastobeatleast65%fullinordertodepart for its destination. Otherwise it waits until this rate is achieved. The maximum waiting time is the other policy related to customer service.Thiswaitingtimeisrestrictedtobeat most 3 days to provide a good service to distributors. After 3 days, even if a truck is not65%full,itwillleavethewarehouseeither by truck or by cargo, whichever is more efficient. MTM does not optimize truck loading.SinceMTMdoesnotpla ninsidethe truckaloadingproblemmayoccur.Giventhe differenceinshapeofthevariousgoodsbeing transported, not all items planned by MTM may be loaded on a truck due to space constraints. As a result, volumes were increasedtoenablethefeasibilityoftheplans generated by the software. While truck load optimization would be feasible for simple deliveries between two points, the Vestel distribution problem is significantly more complex due to routes that have multiple drop‐off points. As a result, the planning objective is not to find the loading that maximizes truck utilization, but rather the loadingthatallows forthebestunloadingof LutfuSagbansua JournalofQafqazUniversity 118 trucks without having to load and unload differentitemsatthevariousdrop‐offpoints. In 2002, Vestel scheduled on average 125 truckseverydayanddelivered49,000products to1000differentlocationseverymonthusing thisplanningsystem. Table2.TransportationFigures Year Month Amount Total Scheduled Truck Volume (dm 3 ) Cumula‐ tive Truck Utiliza‐ tion January 41,153 18,667,200 61% February 43,160 16,691,200 57% March 35,594 17,062,400 57% April 46,284 25,747,200 68% May 58,658 32,291,200 64% June 64,319 25,102,400 72% July 60,552 35,147,200 70% August 46,983 26,148,800 82% September 43,418 20,894,731 85% October 52,533 26,940,860 73% November 69,612 32,733,792 66% 2002 December 68,257 25,111,986 76% January 77,063 28,046,400 89% February 82,877 25,745,600 91% March 104,717 33,944,000 90% April 115,406 31,480,000 95% May 158,242 43,228,800 93% 2003 June 154,923 42,427,200 90% IMPLEMENTATIONISSUESFORVESTEL Theresultsobtainedfromtheimplementation of Manugistics were phenomenal. The truck utilization went up while the transportation costsdecreasedbetween1999‐2003. Table3.DecreaseinTotalTransportationCostfrom 1999to2003 1999 2000 2001 2002 2003 IndexTrans. Cost/Sales Revenue 100.00 119.92 96.69 84.08 81.32 IndexTrans. Cost/Costof GoodsSold 100.00 118.21 98.84 87.91 80.67 Indexof TL/dm3 transporta‐ tion 100.00 109.82 124.61 158.93 163.12 In 2002, transportation costs were decreased by 46% despite the increase in diesel prices and increase in Consumer Price Index. The unit cost of transportation per item went downinsomecasesbyasmuchas75%. Table4.TheUnitTransportationCostDecrease Between1999‐2002 Products %ChangeinUSD TV‐42.92% WashingMachine‐47.36% Refrigerator‐12.92% DishWasher‐44.73% MiniMusicPlayer(portable)‐68.62% Midimusicplayer‐41.50% Micromusicplayer‐28.64% Smallhomeappliances‐51.25% Receiver‐66.41% Dishantenna‐33.52% TVrack‐62.49% Minirefrigerator‐48.97% Carpetwashingmachine‐29.31% Airconditioner(split)‐76.69% Airconditioner(window)‐64.48% Computer ‐66.20% Aspirator ‐52.06% Oven ‐60.77% Stove ‐34.78% Flasheater‐45.10% Vacuumcleaner‐71.20% In addition to the new planning system, a numberofotherfactorswerealsoinstrumental in achieving high utilization rates. First, the number of orders entered manually into the system decreased. The total volume also increasedin2003. Increaseinpre‐paidord ershelpedtoachieve amoreevendistributionof theorderswithin amonth. DecisionSupportSystems:ACaseStudyinVestelDurableGoodsMarketing Number18,2006 119 Table 5. Weekly Distribution of the Monthly Revenue and Truck Utilizations FirstWeek SecondWeek ThirdWeek FourthWeek 2003Jan‐June 19.5% 20.8% 24.3% 35.4% 2002Jan‐June 10.28% 17.17% 19.91% 52.65% WeeklyDistribution 2002Jan‐Dec 9.92% 18.42% 19.62% 52.04% 2003Jan‐June 93% 94% 71% 90% 2002Jan‐June 41% 44% 53% 79% CumulativeTruck Utilization 2002Jan‐Dec 78% 72% 75% 78% Figuresbelowreflecttheincreasedtruckutilizationratesandthetotalscheduledtruckvolumes. Truckutilizationratesarecalculatedusingthefollowingformula:CumulativeTruckUtilization= TotalTransportedVolume(dm 3 )/TotalScheduledTruckVolume(dm 3 ). Figure 1. Cumulative Truck Utilization (%) Cumulative Truck Utilization 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% January February March April May June July August September October November December January February March April May June 2002 2003 Cumulative Truck Utilization Figure 2. Total Scheduled Truck Volume (dm 3 ) Total Scheduled Truck Volume (dm3) 0 5000000 10000000 15000000 20000000 25000000 30000000 35000000 40000000 45000000 50000000 January February March April May June July August September October November December January February March April May June 2002 2003 Total Scheduled Truck Volume (dm3) LutfuSagbansua JournalofQafqazUniversity 120 CONCLUSIONS Decision support systems for supply chain managementareafastgrowingsector of the logisticssoftwareindustry.DSSswillcontinue evolvingandadoptingstandardfeaturesand interfacesinordertoadapttothecompetitive environmentandprovidetheflexiblesolutions requiredintoday’smarkets().Sincethebasic data that are required to make decisions are being collected, there is a strong drive to utilizethisinformationinsophisticatedways to gain competitive advantage by improving service and cutting supply chain costs. ‘Integration with ERP systems’, ‘Improved optimization’,and‘Developmentofstandards’ are the current major trends in DSS and especially supply chain DSS and advanced planningsystems. ThesuccessthatVestelhasexperiencedatthe endoftheimplementationofaDSSmodelin the distribution planning process has once againprovedtheimportanceandvitalroleof DSSineffectivesupplychainpractices. REFERENCES Bechtel, C., Jayaram, J., (1997), “Supply Chain Management: A strategic Perspective”, The InternationalJournalofLogisticsManagement8,pp. 15‐34. Christopher,M.,(1998),“LogisticsandSupplyChain Management. Strategies for Reducing Cost and Improving Service”, 2 nd ed., Financial Times Professional,London,UK. Deutsch, C. H., (1996), “New Software Manages SupplytoMatchDemand”,NewYorkTimes. Feigin, G., An, C., Conners, D., Crawford, I., (1996), “ShapeUp,ShipOut”,OR/MSToday,24‐30. Fox, M. S., Chionglo, J.F., Barbuceanu, M., (1993), “TheIntegratedSupplyChainManagementSystem” Workingpaper,UniversityofToronto. Hewitt, F., (1994), “Supply Chain Redesign”, The InternationalJournalofLogisticsManagement,5,pp. 1‐9. Lee, H. L., Billington, C., (1992), “Managing Supply Chain Inventory: Pitfalls and Opportunities”, Sloan ManagementReview,33(3),65‐73. Proudlove, N. C., Vadera, S. K., Kobbacy, A. H., (1998), “Intelligent Management Systems in Operations”,TheJournaloftheOperationalResearch Society,Vol.49,No.7,pp.682‐699. Ross, D. F., (1998), “Competing Through SupplY Chain Management. Creating Market‐Winning Strategies Through Supply Chain Partnerships”, KluwerAcademicPublishers,Boston,MA,etal. Simchi‐LeviD.,Kaminsky,P.,Simchi‐Levi,E.,(2000), Designing and Managing the Supply Chain, McGraw‐Hill,USA. Swamithan, J. M., (1996), “Quantitative Analysis of Emerging Practices in Supply Chains”, Ph.D. Dissertation,GSIA,CarnegieMellonUniversity. Swaminathan,J.M.,Smith,S.F.,Sadeh,N.M.,(1996), “Modeling Supply Chain Dynamics: A Multiagent Approach”,DecisionSciences. Tan, B., Aksin, Z., (2004), “Vestel: Improving Distribution Management and 3PL Relations”, Case StudyforKocUniversity. Vierck, K.R., (1981), “Decision Support Systems: An MISManagerPerspective”,MISQuarterly,pp.35‐48.