CONCURRENT AIRCRAFT ROUTING AND MAINTENANCE SCHEDULING

Transcription

1 JOURNAL OF AERONAUTICS AND SPACE TECHNOLOGIES JANUARY 2011 VOLUME 5 NUMBER 1 (73-79) CONCURRENT AIRCRAFT ROUTING AND MAINTENANCE SCHEDULING Ilkay ORHAN * Anadolu Unversty School of Cvl Avaton PO Bo Esksehr, Turkey orhan@anadolu.edu.tr Muzaffer KAPANOGLU Esksehr Osmangaz Unversty Faculty of Engneerng Industral Engneerng Department Meselk, Esksehr, Turkey muzaffer@ogu.edu.tr T. Hkmet KARAKOC Anadolu Unversty School of Cvl Avaton PO Bo Esksehr, Turkey hkarakoc@anadolu.edu.tr Receved: 12 th August 2010, Accepted: 21 st January 2011 ABSTRACT The arcraft routng s the process of assgnng each ndvdual arcraft wthn each fleet to flght legs. The Federal Avaton Rules requre the mantenance of all the arcrafts after specfed hours of perod as mandatory. The mnmum total mantenance cost s provded as a result of the lost flght tme whch s brought to mnmum. The common polces n ths busness sector follow the practces the mantenance of an arcraft once n 3-4 days perodcally. Ths polcy mnmzes the rsk of groundng of arcraft n the cost of the lost flght hours. In ths study, we propose a concurrent, mathematcal modelng approach for daly flght route and mantenance schedulng based on recorded flght hours. The model has been solved usng CPLEX/GAMS MILP Software. The proposed approach was appled to the daly flght route-mantenance schedule problems of the domestc flghts of two companes. Keywords: Arcraft Routng, Arcraft Mantenance Plannng, Ar Transportaton, Integer Programmng. BÜTÜNLEŞİK UÇAK ROTALAMA VE BAKIM ÇİZELGELEME ÖZET Uçak rotalama, her br flo çersndek her uçağın uçuş bacaklarına atanma sürecdr. Ulusal Havacılık Kuralları, bell br uçuş saatnden sonra bütün uçaklara zorunlu olarak bakım uygulanmasını gerektrr. En düşük toplam bakım malyet, kullanılamayan uçuş zamanı en küçüğe getrlmesyle sağlanır. Havacılık sektöründe zlenen genel poltka her 3-4 günde peryodk olarak uçaklara bakım uygulamaktır. Bu poltka, kayıp uçuş saat malyetnde uçakların yerde kalma rskn en küçüklemektedr. Bu çalışmada, uçuş saatler temel alınarak günlük uçuş ve rota çzelgeleme çn bütünleşk matematksel modelleme yaklaşımı önerlmştr. Model, CPLEX/GAMS MILP yazılımı kullanılarak çözülmüştür. Önerlen yaklaşım ç hat uçuş gerçekleştren k şrketn günlük uçuş rota bakım çzelgeleme problemlerne uygulanmıştır. Anahtar Kelmeler: Uçak Rotalama, Uçak Bakım Çzelgeleme, Hava Ulaştırma, Tamsayı Programlama. 1. INTRODUCTION Ar travel has been one of the bggest developng sectors n the global ndustry snce the frst flght made by Orvlle and Wlbur Wrght n Arlnes carred appromately 2.25 bllon people n They have made a global proft of $5.6 bllon out of an ncome of $490 bllon [1]. In the years between 2007 and 2026, t s epected that there wll be an ncrease of 4.7% n the passenger traffc and an ncrease of 4.7% n the cargo traffc [2]. Whle the percentage of ncrease n the World Cvl Avaton Sector was 5% n the last few years, t was 53.3% n Turkey [3]. In comparson to 2002, the yearly passenger numbers had ncreased by 350% n the domestc routes and by 63.2% n the nternatonal routes [3]. It has been estmated that the Turksh Cvl Avaton Sector wll ncrease by 4.5%-5% annually and wll buy about 250 arcrafts worth $ 21 bllon [4,5]. * Correspondng Author 73

2 Because of very hgh cost of arcraft purchase and management n the arlne sector, the managements are tryng to brng down the cost n the present compettve envronment. As the arcraft purchase cost cannot be reduced, the managements mostly concentrate on ts operaton cost. The arlne management cannot control the prce of fuel and the rent of arport area drectly. The only cost that can be controlled s the mantenance cost of the arcraft. The management that can plan the arcraft mantenances optmally can make mllons of dollars worth of savngs, and hence can be better placed over ts rvals. The arlne managements prepare ther mantenance programs by takng the documents of MPD (Mantenance Plannng Documents), AMM (Arcraft Mantenance Manual), CMM (Component Mantenance Manual), SBs (Servce Bulletns) and SLs (Servce Letters) as reference [6]. The prepared programs are put nto applcaton f found suffcent by The Cvl Avaton Authorty of the country concerned. The arcraft mantenance perod used, are appled accordng to A, B, C and D codes of the check perods [7, 8]. The parts and components on the arcraft are assgned to the letter coded control packages and the relevant mantenance s done lke ths. Moreover, along wth ths, today for the latest verson arcrafts lke B /700/800/900, the alphabet coded mantenance applcatons are not now suggested by the frm n the mantenance plannng documents as reference [9]. The producer companes gve the mantenance perods, arcraft flght tme, arcraft landng numbers, and year-lfe lmt types of measurement to the users as reference. The arlne management can get the parts and components whch form the most of arcraft mantenance package done n a better perod by takng nto account the flght tme. As a concluson, the mantenance applcaton can be done by takng nto account the real flght tmes rather than takng the daly average flght perod of the arcrafts. Thus, less arcraft mantenance can be acheved annually by applyng mantenance accordng to the lmt value or close to the value to rato of arcraft usage. For eample, f the arcraft, whch mantenance nterval s 400 flght hours, s done mantenance at around 400 flght hours nstead of 360 flght hours, the eleventh mantenance cost can be avoded. On top of ths mantenance cost savngs, the arcraft utlzaton wll be mproved due to shorter groundng tme. Federal Avaton Admnstraton n the Unted States requres the letter coded check packages. These checks vary ther scope, duraton, and frequency. A- type checks have to be performed every 65 flght hours. The other checks have longer duratons. The type of B check s performed every 300 to 600 hours of flght. The C and D checks are performed every one to four years. The arlnes n practce are performed A type checks at most 35 to 40 hours of flyng. The tme requred to perform A checks s somewhere between 3 and 10 hours [10, 11]. Naturally, A-checks can be held accountable for most of the mantenance operatons. In the lterature, most of the studes assgn arcrafts to the routes followng an A-check. Snce each arcraft practcally spends around 12 hours per day, there remans three to four sold days before the net A-check. Therefore, an A- check s supposed to be ssued every four or fve days. Havng consdered the fact that legally allowed tme between A-checks s 65 hours, ths practcal approach results n the loss of 17 to 30 hours of flght tme accountng roughly for 25% to 45% of the duraton. The operatons management takes the advantage of ths so-called cushon tme to prevent any severe penalty for mssng an A-check under the assumpton of daly plans. If the plans could be performed and eecuted on hourly bass wthn a robust plannng method, these steady loss percentages are epected to be reduced sgnfcantly wthout any mssng of checks. These savngs would be much more tangble n cases of route changes due to bad weather condton, llness of crew or techncal reasons. In ths study, a concurrent daly arcraft mantenance and route plannng approach has been proposed to mnmze arcraft mantenance costs whch form 18-23% of the arlnes drect operatng cost [12]. The proposed method ncorporates the arcraft routng and mantenance schedulng problems nto a mathematcal model based on a flght-hour trackng approach. The model proposed has been solved usng CPLEX/GAMS MILP Software. The approach has also been appled to the daly flght route-mantenance schedule problems of the domestc flghts of two companes. 2. LITERATURE REVIEW Daskn et al. has shown that by puttng the arcraft routes n the sngle hub and spoke network structure, the management proft can be ncreased. The problem was modeled as MILP and was solved wth Lagrangan relaaton [13]. Feo and Bard had presented a flght program development model whch could be placed n the mantenance centers and whch could solve the A type check requrement n a better way. The problem was formulated for Amercan Arlnes as a mnmum cost, mult commodty flow network wth ntegral constrants. For the lnear programmng relaaton soluton and as t was qute bg, a two-stage heurstc was used [14]. Kabban and Patty the arcraft routng problem for Amercan Arlnes had been formulated n such a way that every column represented probable weekly arcraft routes and the rows represented the flghts. In ths study, the mantenance has been performed once n every three days [15]. Hane et al. had modeled the fleet assgnment problems wth sde lmtatons as a large mult-commodty network flow wth sde constrants 74

3 defned on a tme-epanded network. The method used to solve problem was the nteror-pont algorthm and branchng. Moreover, the authors had used some preprocessng method n order to dmnsh the dmenson of problem [16]. Clarke et al. etended the study n Hane et al. [16] addng to the model mantenance and crew constrants. The soluton that cannot provde the requrement of mantenance was accepted to be napproprate. The requrement for mantenance had been classfed as short term and long term accordng to the perod t was performed. In the artcle, how mantenance constrant and crew assgned to the fleet had got ntegrated s not mentoned clearly [17]. Clarke et al. modeled the arcraft rotaton problem as an Euleran tour wth sde constrant that mamzed the through value of connectng flghts and satsfes mantenance constrants. The model was smplfed wth preprocess technque. It was solved wth Lagrangan relaaton and subgradent optmzaton [10]. Desaulners et al. had analyzed the mamum proft from daly heterogeneous arcraft routng and schedulng problem and the two models were presented. The lnear relaaton of frst model was solved by usng column generaton technque; the lnear relaaton of the second model was solved by usng Dantzg Wolfe decomposton approach [18]. Gopalan and Tallur modeled mantenance routng problem and had used polynomal-tme algorthm for the three-day mantenance and the balance-check vst requrements. In the model, swappng was done n the flghts as well as fleet to obtan approprate arcraft mantenance [11]. Barnhart et al. presented a strngbased model and used a branch-and-prce approach to solve combned fleet assgnment and arcraft routng problem. The dsadvantage of ths model was havng mllons of seres by flght programs havng hundreds of flghts [19]. Cordeau et al. modeled the arcraft routng and crew schedulng problems smultaneously, and used Bender decomposton approach for soluton [20]. Srram and Haghan analyzed the mantenance schedulng and arcraft reassgnment. The authors took nto consderaton both A and B type checks n the formulaton and used a heurstc method that was a hybrd of random search and depth frst search [21]. Sarac et al. solved the arcraft mantenance routng problem by takng nto consderaton the mantenance slots and the man-hours whch can be used at the mantenance staton [22]. 3. THE FORMULATION OF PROBLEM The flght network s modeled as a connecton network n whch nodes represent flght legs and arcs represent feasble connectons between the flght legs. In the connecton network, the arc (, j) ests f the flght legs and j are beng realzed by the same arcraft. The destnaton cty of node j and arrval cty of must be the same. It also requres that the arrval tme of flght plus the turn tme s less than or equal to the departure tme of flght leg j. In the model, the followng notaton s used. and j ndces for flght legs (nodes) k nde for arcraft o dummy source node connected to every node N t dummy destnaton node connected to every node N c nde for cty (arport) c C mt mantenance type K set of arcraft N set of staton C set of cty (arport) T d departure tme of flght leg T j duraton of flght leg T s requred tme for preparng the arcraft for net flght after flght leg M a consderable bg number Tdo ok departure tme for arcraft k at dummy source o Tdt tk departure tme for arcraft k at dummy snk t 1, f arcraft k fles flght leg and j consequently; 0, otherwse Ca c 1, f arrval cty of flght leg s c; 0, otherwse Cd jc 1, f departure cty of flght leg j s c; 0, otherwse Cmt mt,c 1, f overnght arport c can serve for mantenance type mt; 0, otherwse Amt k 1, f arcraft k needs mantenance type mt; 0, otherwse. r k legal remanng flyng hours for arcraft k. DFFc() the flght begnnng n cty c OWNc(k) the arcraft k spent the nght n cty c KM(k) set of arcrafts needs mantenance KMw(k) set of arcrafts needs mantenance performed accordng to day perod KMc() the flght performed to mantenance cty Nth_u the set of flght legs performed as a through flght Wth ths notaton, the problem formulaton becomes 1 Mn ( rk ) ( ) Tj (1) ( rk ) N = 1 / o, t (2) N { } = 1 k K ve = o (3) = 1 k K ve j = t (4) { o t} = 0 /, vek K (5) jk = 0 (6) ojk N tk 75

4 Td + T + Ts Td j M(1 { o, t} K ve (, j) / Tdook + To Td M (1 K ve ( ) t K ve j tk ( ) ok Td + T + Ts Tdt M(1 ) ojk c C Ca Cd c jc ) ) tk ve j (7) (8) (9) (10) Cd Ca j ve k K (11) j DFFc( j) jc tk N m Mc C N oc = 1 OWNc( k) ve = o T (12) Ca Cmt, A, = 1 KM (13) j KMc( ) c r k mt c mt k K (14) = 1 KMw ve j = t (15) Nth_u th_u t oj Tdo = 1 (16) = 0 k K (17) ok 0 ve Tdt 0 K (18) tk {,1} K ve j 0, (19) The objectve functon mnmzes the total remanng legal flght tme by gvng hgher prortes to the arcrafts wth lower remanng flght tmes. The proposed objectve functon ncreases the utlzaton of arcrafts yeldng savngs over mantenance costs. The constrants for flght leg coverage (2) guarantee that an arcraft wll be assgned to each flght leg. The arcraft coverage conssts of two set of constrants. The frst one (3) assgns every arcraft to the begnnng of a flght leg, whle the second one (4) ensures that every arcraft whose flght ends at a dummy snk wll be assgned to a new flght for the day after. The flght conservaton constrants set (5, 6) mantan the balance between the number of connecton arcs comng and leavng node. The flght conservaton constrant (5) enforces each arcraft to make a crculaton n the flght schedule. Thus, the arcraft contnues flghts untl the flght reachng to a dummy pont. The flght conservaton constrant (6) s used for equalng the number of connecton arcs leavng a dummy source node wth the number of flght connectons comng to a dummy snk. Dong so, the model provdes the same number of arcraft at every arport for repeatng flghts. The sequence flght legs are provded by feasble tme connectons and feasble cty-par connectons. The frst one (7) ensures feasble connecton wth takng care of the tme between flghts. If the sum of the duraton of flght tme and the turn tme at the end of a flght s less than the tme of the begnnng of the net flght, ths constrant connects the flght and j each other wth flght connecton arc. The departure tme constrant n the same way s wrtten for dummy source node (8) as well as for dummy snk (9). The constrant set (10) provdes feasble connecton by takng nto consderaton of arports for the flghts taken place wthn the day. In ths constrant, for arcraft k the varable s 1 only when the departure cty of flght leg j s the same wth the arrval cty of flght leg. Otherwse, all the varables are equal to zero. The constrant set (11) s a constrant for dummy source node. It provdes that every arcraft s assgned a startng flght leg. The constrant set (12) s the group whch provdes the assgnment of frst flghts from the arports by takng nto consderaton the arport the arcrafts spend the nght. The constrant set (13) provdes that the arcrafts havng value below certan lmts are assgned to the arports for mantenance n ther last flght after they completed approprate routes. Ths set of restrctons ensures that the arcrafts are routed to the arports where the requred mantenance can be performed. The parameter that represent the mantenance lmts can be determned by the decson maker wthn the range of 9-12 flght hours accordng to average daly flght tme. The arcrafts wth flght hours eceedng the lmtng range mght overnght at the arport of the last flght. The constrant group (14) s used to make sure that the assgned routes reman wthn the legal flght tme. The local Cvl Avaton Authorty ssues severe fnes ncludng revokng flght lcenses to arlnes companes whch neglect the mantenance wthn legally allowed flght perod tme n accordance wth the relevant regulatons. Although most of the arcraft mantenance procedures take the flght hours as bass, there may stll be some day-based mantenance operatons mposed by the manufacturers. The constrant (15) provdes routes n such a way that at the end of n days the arcrafts are assgned to ctes wth approprate mantenance centers. The constrants of (16) are formed to buld through flghts nstead of connectng flghts, snce arlnes provdng ths servce to the passengers reported to have ncreased ther proft, and have become preferred arlnes [23]. The major advantage of a through flght over a connectng flght s to save passengers from the hassle of movng from one arcraft to another. Nth_u nde set for these constrants represents the set of through flghts. The duraton of flght from dummy source to any arport s zero. The departure tmes of the dummy source and dummy snk for each arcraft k s dfferent from zero. 4. APPLICATIONS Three sets of flght programs for eperments have been prepared based on the data of two commercal arlnes. The sets nclude 22, 64, and 136 flghts consecutvely. The problems have been solved by usng CPLEX/GAMS med nteger lnear program solvng software. The results of the eperments obtaned have been presented and nterpreted. 76

5 In the problem set of 22 flghts, the number of arcrafts s set to four; and the number of through flghts s set to 1. The results of concurrent arcraft route and mantenance plannng model, are shown n Fgure 1 for each arcraft. The fgure ndcates that the arcrafts wth longer legal remanng flyng hours are utlzed less whle the ones wth shorter remanng flyng hours are utlzed up to 100%. For nstance, the arcraft havng number 1 on the tal has the hghest remanng flyng hours (850 mnutes) wth the lowest utlzaton percentage (42%). The arcraft havng number 3 on the tal havng 450 flght mnutes were utlzed 100%. As seen n Fgure 1, the model assgned the arcrafts wth mnmum remanng legal flght hours to routes. However, ths does not mean that an ascendng sortaton of remanng flght hours can be used to solve as a prortzaton scheme n assgnng arcrafts to routes. It depends on the place where the arcraft s, the flght tme, other connectng and through flghts, and the arport t has landed n the last flght. The sample data sheet for 22 flghts s gven n Table 1, and the routes map s shown n Fgure 2. In the set of 64 flght-problems, the number of arcrafts s set to 11. The flght program was formed n terms of hub-and-spoke network wth a sngle hub. Ths means that there was less connecton possblty among these flghts. The total tme perod of 64 flght perods-11 arcrafts assgned to mantenance and route plannng and ther legal remanng flyng hours comparson have been shown n Fgure 3. The sample data sheet for 64 flghts s gven n Table 2, and the routes map s shown n Fgure 4. Table 2. Sample Data Sheet for 64 Flghts-Test Problems. Fgure 1. For 22 Flghts Concurrent Arcraft Routng and Mantenance Schedulng Solutons. Table 1. Sample Data Sheet for 22 Flghts-Test Problems. Fgure 3. For 64 Flghts Concurrent Arcraft Routng and Mantenance Schedulng Solutons. Fgure 2. For 22 Flghts Routes Map. Fgure 4. For 64 Flghts Routes Map. 77

6 In the thrd problem consstng of 136 flghts we focused on the computaton tmes over dfferent arcraft numbers based on the proposed approach. Ths problem s a hub-and-spoke network wth two hubs. The model has been solved for varyng number arcrafts from 30 to 38. Snce ncreasng number of arcrafts generates more alternatves, the ncrements n the computaton tme are not steep resemblng a lnear trend. For 136 flghts, daly concurrent arcraft mantenance and routng model, soluton and tme relatonshp wth dfferent numbers of arcraft have been shown n Fgure 5. The sample data sheet for 136 flghts s gven n Table 3, and the routes map s shown n Fgure 6. Fgure 5. For 136 Flghts The Soluton Tmes for Concurrent Arcraft Routng and Mantenance Schedulng Model Accordng To Arcraft Numbers. Table 3. Sample Data Sheet for 136 Flghts-Test Problems. Fgure 6. For 136 Flghts Routes Map. 5.CONCLUSIONS Wth the developng economy and the lftng of lmtaton n the aerospace sector, the Turksh Cvl Avaton Sector s epected to develop by 4.9% n the comng years. In arlnes ndustry where competton s very stff, a successful management must have a soluton plan to reduce mantenance costs accountng for 18-23% of the total operaton costs. In ths study, a mathematcal-model drven approach s proposed to mamze the flght tmes between mantenances. The model assgn arcrafts to flghts to buld a route by takng the remanng flght tmes nto consderaton as well as the locaton nformaton of the arports that can serve the arcrafts for that specfc mantenance. The model s formulated to address the ssue of hour-based flght tme nstead of the daybased flght tmes of the common practce. Mamzng the flght tmes and assgnng the arcraft to the arports wth the approprate mantenance centers as the fnal destnaton before mantenance perod ends enable the proposed approach to mnmze early and tardy mantenances. Mnmzng early and tardy mantenances promses sgnfcant savngs for arlnes. Another advantage of the approach s to provde smooth mantenance requrements by gvng hgher prorty to the arcrafts wth less remanng flght tmes. Ths prortzaton scheme reduces the peaks and, therefore the tenson over the mantenance centers due to capacty stretches. Instantaneous capacty ncreases requred from mantenance centers usually causes an addtonal cost. Ths cost does not only nclude drect mantenance costs due to etra shfts but also ndrect costs of lower utlzed arcrafts due to nevtable groundng for mantenance actvtes. The eperments have been performed for A-checks. However, once ths approach has been adopted as a concurrent routng-mantenance plannng tool, the savngs wll preval to the other types of checks, B, C and D n the same way. Consecutvely, the arlne operators wll have advantage over ther compettors wthout any volaton of the rules lad down by the cvl avaton authorty. 6. REFERENCES [1] IATA, IATA 2008 Annual Report, Internatonal Ar Transport Assocaton, [2] Arbus, Arbus Global Market Forecast : Arbus, [3] 08turkey.pdf [4] & [5] and-celebrates-growth-n-the-turksh-avaton-market [6] Knnson, H. A., Avaton Mantenance Management, McGraw-Hll, New-York, [7] Orhan, I., Kapanoglu, M., Karakoc, T.H. 2007, Flght-Hour Based Optmzaton for Arcraft Schedulng, INFORMS, 369, WD43, A.B.D. 78

7 [8] Orhan, I., Kapanoglu, M., Karakoc, T.H., 2010, Plannng And Schedulng Of Arlne Operatons, Pamukkale Unversty Journal of Engneerng Scences, 16(2), , [9] Boeng, Secton 1-System Mantenance Program, /700/800/900 Mantenance Plannng Document. s.1.0-5, [10] Clarke, L.W., Johnson, E.L., Nemhauser, G.L., Zhu Z., The Arcraft Rotaton Problem, Annals of Operatons Research, 69, 33-46, [11] Gopalan, R., K. Tallur, The Arcraft Mantenance Routng Problem, Operatons Research, 46, , [12] Radnot, G., Proft Strateges for Ar Transportaton, McGraw-Hll, New York, [13] Daskn, M.S. and Panayotopoulos N.C., A Lagrangan Relaaton Approach to Assgnng Arcraft to routes n Hub and Spoke Networks, Transportaton Scence, 23(2), 91-99, [14] Feo, T.A., Bard J. F., Flght Schedulng and Mantenance Base Plannng, Management Scence, 35, , [15] Kabban, N.M. and Patty B.W., Arcraft Routng at Amercan Arlnes, Proceedngs of the Thrty-Second Annual Symposum of the Arlne Group of the Internatonal Federaton of Operatonal Research Socetes, Budapest, Hungary, [16] Hane, C.A., Barnhart, C., Johnson E.L., Marsten R.E., Nemhauser G.L., Sgsmnd G., The Fleet Assgnment Problem: Solvng A Large Scale Integer Program, Mathematcal Programmng, 70, , [17] Clarke, L.W., Hane, C.A., Johnson, E.L., Nemhauser, G.L., Mantenance and Crew Consderatons n Fleet Assgnment, Transportaton Scence, 30, , 1996., [18] Desaulners, G., Desrosers, J., Dumas, Y., Daly Arcraft Routng and Schedulng, Management Scence, 43, , [19] Barnhart, C., Boland, N.L., Clarke L.W., Johnson E.L., Nemhauser, G.L., Flght Strng Models for Arcraft Fleetng and Routng, Transportaton Scence, 32(3), , [20] Cordeau, J.F., Stojkov c, G., Soums F., Desrosers, J., Benders Decomposton for Smultaneous Arcraft Routng and Crew Schedulng, Transportaton Scence, 35, , [21] Srram, C. and Haghan, A., An Optmzaton Model for Arcraft Mantenance Schedulng and Re-Assgnment, Transportaton Scence, 37, 29-48, [22] Sarac, A., Batta, R., Rump, C.M., A Branchand-Prce Approach For Operatonal Arcraft Mantenance Routng, European Journal of Operatonal Research, 175, , [23] Ahuja, R.K., Goodsten, J., Mukherjee, A., Orln, J.B., Sharma, D., A Very Large-Scale Neghborhood Search Algorthm for the Combned Through and Fleet Assgnment Model, Workng paper , MIT Sloan School of Management, VITAE Asst. Prof. Dr. Ilkay ORHAN He s an Assstant Professor n School of Cvl Avaton at Anadolu Unversty, Esksehr, Turkey. Dr. Orhan also has the appontment as deputy drector of the Cvl Avaton Research and Applcaton Center. He receved hs MSc and Ph.D n 2003 and 2007, respectvely. Hs research nterest ncludes arlne plannng and schedulng problems. Assoc. Prof. Dr. Muzaffer KAPANOGLU Dr. Kapanoglu s an Assocate Professor n Esksehr Osmangaz Unversty. He also has the appontment as the charman of the Department of Informatcs n the same unversty. He receved hs undergraduate and Master of Scence degrees n the Department of Industral Engneerng, Anadolu Unversty, Esksehr, Turkey. He got hs Ph.D. degree from the College of Engneerng, Unversty of South Florda, n the Department of Industral Engneerng n He teaches Decson Support Systems, Operatons Research and Heurstc Optmzaton. He also has teachng appontments wth Baskent Unversty, Ankara for over a decade. Among hs research areas are large-scale plannng and schedulng for proft and non-proft enterprses wth machne learnng and heurstc methods. Prof. Dr. T. Hkmet KARAKOC He s a Full Professor n School of Cvl Avaton at Anadolu Unversty, Esksehr. He receved hs MSc and Ph.D n 1983 and 1987, respectvely. He has publshed many papers at varous natonal and nternatonal conferences, whle he has authored several books. He s also a member of many journals and assocatons. Hs research areas nclude avaton, energy and energy economy, gas turbne engnes, fuels, solaton and nstallaton. Mr. Karakoc s drector of the Cvl Avaton Research and Applcaton Center at Anadolu Unversty. 79