Wong Chi Fat. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2002. / Includes bibliographical references (leaves 107-115). / Abstracts in English and Chinese. / List of Tables --- p.ix / List of Figures --- p.x / Chapter 1. --- Introduction --- p.1 / Chapter 1.1 --- Motivation --- p.1 / Chapter 1.2 --- Literature Review --- p.3 / Chapter 1.2.1 --- Shortest Path Problem --- p.4 / Chapter 1.2.2 --- Vehicle Routing Problem with Time Windows --- p.6 / Chapter 1.3 --- Thesis Outline --- p.9 / Chapter 2. --- Time-varying Shortest Path with Constraints in a 2-level Network --- p.11 / Chapter 2.1 --- Introduction --- p.11 / Chapter 2.2 --- Problem Formulation of TCSP --- p.12 / Chapter 2.3 --- Arbitrary Waiting Time --- p.13 / Chapter 2.4 --- TCSP in a 2-level Network --- p.15 / Chapter 2.4.1 --- Problem Formulation of TCSP in a 2-level Network --- p.17 / Chapter 2.5 --- Algorithms Solving TCSP in a 2-level Network --- p.20 / Chapter 2.5.1 --- Exact Algorithm --- p.21 / Chapter 2.5.2 --- Heuristic Algorithm --- p.23 / Chapter 2.6 --- Concluding Remarks --- p.30 / Chapter 3. --- Vehicle Routing Problem with Time Windows and Stochastic Travel Times --- p.32 / Chapter 3.1 --- Introduction --- p.32 / Chapter 3.2 --- Problem Formulation --- p.34 / Chapter 3.3 --- General Branch-and-cut Algorithm --- p.42 / Chapter 3.4 --- Modified Branch-and-cut Algorithm --- p.44 / Chapter 3.4.1 --- Prefixing --- p.45 / Chapter 3.4.2 --- Directed Partial Path Inequalities --- p.47 / Chapter 3.4.3 --- Exponential Smoothing --- p.50 / Chapter 3.4.4 --- Fast Fathoming --- p.54 / Chapter 3.4.5 --- Modified Branch-and-cut algorithm --- p.56 / Chapter 3.5 --- Computational Analysis --- p.57 / Chapter 3.5.1 --- "Performance of Prefixing, Direct Partial Path Inequalities and Exponential Smoothing" --- p.57 / Chapter 3.5.2 --- Performance of Fast Fathoming --- p.63 / Chapter 3.5.3 --- Summary of Computational Analysis --- p.67 / Chapter 3.6 --- Concluding Remarks --- p.67 / Chapter 4. --- System Features and Implementation --- p.69 / Chapter 4.1 --- Introduction --- p.59 / Chapter 4.2 --- System Features --- p.70 / Chapter 4.2.1 --- Map-based Interface and Network Model --- p.70 / Chapter 4.2.2 --- Database Management and Query --- p.73 / Chapter 4.3 --- Decision Support Tools --- p.75 / Chapter 4.3.1 --- Route Finding --- p.75 / Chapter 4.3.2 --- Delivery Planning --- p.77 / Chapter 4.4 --- System Implementation --- p.80 / Chapter 4.5 --- Further Development --- p.82 / Chapter 5. --- Vehicle Routing Software Survey / Chapter 5.1 --- Introduction --- p.83 / Chapter 5.2 --- Essential Features in CVRS Nowadays --- p.84 / Chapter 5.2.1 --- Common Features --- p.34 / Chapter 5.2.2 --- Advanced Features --- p.90 / Chapter 5.3 --- Concluding Remarks --- p.94 / Chapter 6. --- Summary & Future Work --- p.97 / Appendix A --- p.101 / Appendix B --- p.104 / Bibliography --- p.107

Thesis (PhD)--Stellenbosch University, 2002. / ENGLISH ABSTRACT: The traditional ways in which manufacturers of crop protection products exercised control over distribution systems have diminished in their impact. Control was based on quality products, the power of the brand, an installed customer base and a broad portfolio of mostly patented products. Manufacturers of crop protection products need to review the manner in which these products are marketed, which distribution system management models are to be used, and the importance that is placed on distribution system management in the marketing mix. The primary aim of this study was defined as: "The development of appropriate distribution system management models for application in South Africa and Australia~. The secondary aim was defined as: "Establishing the key factors which determine the relationship between manufacturers and distributorsn . The dominant economic characteristics of the global crop protection industry are deemed to consist of a global market valued at US $28,090 mio, a mature market in a decline phase, the rapid consolidation of industry players, increased generic product manufacturer activity and an overall decline in manufacturer profitability. In South Africa crop protection product manufacturers use approximately 46 independent distributors to market products on farms through 600 affiliated commissioned sales agents. In Australia the distribution of crop protection products is much more concentrated. Distribution is essentially controlled by five national distributors and their coupled salaried representatives. Manufacturers therefore rely on third parties for the marketing of their products to farmers in both countries. The applicable problem statement has been formulated to select a distribution system management model that will: (i) optimally balance direct distribution related cost and subsequent levels of control over distributors; (ii) maximise the probability that a distributor will buy and actively promote the complete product portfolio of a for high levels of interpersonal relationship maintenance; and (vi) manufacturers have to instill the philosophy that distribution system management is part of a manufacturer's strategic business and marketing focus and not simply a task to be performed by a third party. / AFRIKAANSE OPSOMMING: Die tradisionele maniare waarop die vervaardigers van oesbeskermingsprodukte beheer kan uitoefen oor verspreidingsisteme is nie meer so suksesvol nie. Beheer was gebaseer op kwaliteil produkte, die mag van die handelsmerk:n gevestigde klientebasis en 'n wye produktereeks,bestaande uit gepatenteerde produkte, Vervaardigers van oesbeskermingsprodulcte moet die wyse waarop hierdie produkte bemark word, watter verspreidingsisteem bestuursmodelle gebruik word en die belangrikheid wat geplaas word op die bestuur van die verspreidingsisteme in die totala bemarkingspoging in heroorweging neern. Die primere doelstelling van hierdie studie is: "Die ontwikkeling van toepaslike verspreidingsisteem bestuursmodelle vir toe passing in Suid-Afrika en Australia. "Die sekondere doelstelling is: "Die vas stelling van die sleutelfaktore wat die verhouding tussen velVaardigers (verskaffers) en verspreiders bepaar. Die dominante ekonomiese kenmerke van die globale oesbeskermingsindustrie is 'n markwaarde in VSA $ 28,090 miljoen, 'n stagnante mark in 'n agteruitgang lase, die drastiese konsolidasie van die industrie spelers, 'n toename in die aktiwiteite van die generiese produkvervaardigers en 'n algemene daling in die winsgewindheid van die vervaardigers. In Suid-Afrika gebruik die vervaardigers van die oesbeskermingsprodukte ongeveer 46 onafhanklike verspreiders om die produkte op plaasvlak te bemark met behulp van 600 geaffilieerde kommissie agente. Die verspreiding van oesbeskermingsprodukte is baie meer gekonsentreerd in Australie. Verspreiding word in wese beheer deur vyf nasionale verspreiders en hulle spanne van verkoopsteenwoordigers wat 'n salaris verdien. In beide lande moet velVaardigers van oesbeskermingsprodukte staatmaak op onafhanklike instansies vir die bemarking van hul produkte op plaasvlak. Die probleemsteiling is dus die uitdaging om 'n verspreidingsisteem bestuursmodel t9 ontwikkel wat: (i) . n optima Ie balans sal gee tussen direkte verspreidingskoste en gekoppelde vlakke van beheer oor verspreiders; (ii) die waarskynlikheid dat . n verspreider die betrokke produkreeks van die vervaardiger sal aankoop en aktief sal promoveer; (iii) dit kan bereik in . n mark waar verspreiders talle bronne het van soortgelyke produkte teen kompeterende pryse. Gebaseer op die sekondere navorsing wat gedoen is wit dit blyk dat daar 'n algemene konsensus is dat daar n nei9in9 weg is van transaksie spesifieke besigheidsverhoudings na besigheidsverhoudings gebaseer op vennootskap gebaseerde benaderings in die interaksie tussen vervaardigers (verskaffers) en verspreiders. Hierdie tipe besigheidsverhouding word gekenmerk deur hoe vlakke van onderlinge vertroue, wedersydse verbintenis, konflik hanteringsmeganismes en goeie wedersydse kommunikasie. Primere navorsing is in Suid-Afrika (253 respondente) en Australia (180 respondente) gedoen. Implikasies vir die konstruksie van verspreidingsisteem bestuursmodelle wat voortvloei uit hierdie navorsing sluit in die belangrikheid dat on vervaardiger on wye en gesogte produkreeks het, die vereiste dat produkte beproefde effektiwiteit het, produkte kompeterend geprys is, die vereiste vir hoe vlakke van verskaffer opvolgdiens en die vereiste vir goeie interpersoonlike verhouding tussen werknemers van beide die vervaardigers en die verspreiders. Hierdie elemente moet ondermeer gerugsteun word deur goeie kommunikasiesisteme. Die "idea Ie" verspreidingsisteem bestuursmodelle is vervolgens voorgestel vir beide Suid-Afrika en Australie, gebaseer op die bevindinge van ondermeer die sekondere- en primere navorsing. Hierdie mode lie het ten doel om toepaslike strukture en besigheidsbenaderings vir vervaardigers daar te stel om hulle behulpsaam te wees om die geidentifiseerde behoeftes van die verspreiders aan te spreek.

Thesis (MBA)--Stellenbosch University, 2007. / ENGLISH ABSTRACT: HM Timber operates five formal sawmills in South Africa, situated around the towns of Tzaneen, Sabie, Mthatha and Kokstad. The Mills in the Kokstad area experience unacceptable lead times during the months of April/May and November/December, when the demand for timber increases and the availability of transport in the area decreases. This problem is expected to intensify with the addition of HM Timber’s new Plywood plant just outside the town of Kokstad. The plywood also adds another dimension to the distribution problem in that the market generally demands smaller deliveries more often. With only one Distribution Centre in Cape Town, HM Timber is not geared to service this market in the rest of South Africa. The main objective of this research report is to determine and define the key success factors in the transport and distribution of timber products and how these factors can be integrated into various strategic options that will address HM Timber’s current distribution problems. A review of the literature summarises the latest trends in modern supply chain management. Supply chains consist of various businesses that aim to optimise all their operations and supply customers with good quality products and excellent service at affordable prices. Competition in modern day business increasingly occurs at supply chain level and the success of individual organisations therefore depends on the success of the entire supply chain. The focus is thus on collaboration and jointly developing supply chain strategies. The literature review ends with a summary of the most important aspects in the effective and efficient management of Distribution Centres. Analyses of HM Timber’s logistics operations reveal that various factors contribute to their transport problems during the months of April/May and November/December. These include the payment of rates that are not market-related, loading problems at the mills, a lack of coordination and the fact that HM Timber is not a contracted client and therefore not regarded a priority. The analysis also includes a discussion on the Western Cape Distribution Centre and the problems this facility is having with general administration, inventory management and the effective use of IT systems. A range of strategic options is available to HM Timber that will allow the company to improve its overall management of logistics. The options investigated include own operations as well as outsourced 3PL and 4PL solutions. To assist HM Timber in selecting the best logistics option, three models are presented to evaluate the various alternatives and to assist in deciding on the best option. The first model evaluates the costs associated with different transport options, while the second model allows the user to compare the costs of different Distribution Centre options. Finally, a multi-criteria decision model is presented that will assist HM Timber in selecting the best option for each mill and region. With the VISA transport model, Manline is rated the best option for all Weza and Singisi’s deliveries. Manline is also rated the best for Langeni’s long-distance deliveries, while HM Timber’s own fleet is rated the best option for Langeni’s local deliveries. Results from the Distribution Centre (DC) models indicate the cost structures to be sensitive to changes in the premiums achieved and changes in the market e.g. the customer base, product types, and volumes sold. The Johannesburg and Port Elizabeth DCs show acceptable EBIT margins based on the projected sales, but the projected EBIT margins for Durban are less than 2 %, making the Durban DC a less than attractive business proposition. A direct distribution solution is proposed for the KwaZulu-Natal area, whereby multiple deliveries are made directly from the mills as well as the new plywood plant in Kokstad. In conclusion, eight factors are identified as key to the successful transportation and distribution of timber products. HM Timber rates very well against the majority of these factors, but needs to focus on a number of issues to be able to maintain their high levels of customer service. These include investing in a dedicated logistics department, improving internal and external collaboration and coordination efforts and improving the efficiency of a number of operational and information technology systems. / AFRIKAANSE OPSOMMING: HM Timber bedryf vyf formele saagmeulens in Suid-Afrika wat geleë is in die omgewings van Tzaneen, Sabie, Mthatha en Kokstad. Die meulens in die Kokstad omgewing ondervind onaanvaarbare vertragings in aflewerings gedurende April/Mei en November/Desember wanneer die aanvraag na houtprodukte toeneem en die beskikbaarheid van vervoer in die Kokstad area afneem. Hierdie probleem gaan na verwagting groter word namate die laaghoutvolumes toeneem. ‘n Verdere uitdaging in die verspreiding van laaghout is die feit dat kliënte oor die algemeen kleiner hoeveelhede, meer gereeld aankoop. HM Timber bedryf slegs een Verspreidingsentrum in Kaapstad en is dus nie gereed om ander dele van Suid-Afrika te bedien nie. Die hoofdoelwit van hierdie navorsingsverslag is om te bepaal watter faktore noodsaaklik is om suksesvol te wees in die vervoer en verspreiding van houtprodukte. Die doel is verder om te bepaal hoe hierdie faktore in verskillende strategiese opsies geïntegreer kan word ten einde HM Timber se huidige verspreidingsprobleme aan te spreek. ‘n Studie van die literatuur gee ‘n opsomming van die nuutste tendense in die bestuur van moderne verskaffingskettings. Verskaffingskettings bestaan uit verskeie besighede wat ten doel het om al hul aktiwiteite te optimiseer en teen bekostigbare pryse uitstekende diens aan hulle kliënte te lewer en produkte van hoë gehalte te voorsien. Kompetisie in die moderne besigheid vind meer en meer plaas tussen die verskillende verskaffingskettings en die sukses van enige individule organisasie is dus afhanklik van die sukses van die algehele verskaffingsketting. Die fokus is dus op samewerking en om gesamentlik strategieë vir die verskaffingsketting te ontwikkel. Die literatuurstudie eindig met ‘n opsomming van die mees belangrike aspekte in die effektiewe en doeltreffende bestuur van verspreidingsentrums. Ontleding van HM Timber se logistieke aktiwiteite toon dat verskeie faktore verantwoordelik is vir die maatskappy se vervoerprobleme tydens April/Mei en November/Desember. Dit sluit in die betaling van nie-markverwante tariewe, laaiprobleme by die meulens, ‘n tekort aan koördinasie en die feit dat HM Timber nie ‘n kontrak met enige vervoermaatskappy het nie. Die analise sluit verder ‘n bespreking van die Verspreidingsentrum in Kaapstad in en die probleme wat hierdie fasiliteit ondervind met algehele administrasie, voorraadbeheer en die doeltreffende gebruik van inligtingstelsels. ‘n Reeks strategiese opsies is beskikbaar wat HM Timber instaat sal stel om algehele logistieke bestuur binne die maatskappy te verbeter. Opsies wat ondersoek word sluit in eie operasies sowel as die uitkontrakteer van die logistieke funksies aan 3PL en 4PL organisasies. Ten einde HM Timber te help om die beste opsie te kies, word drie modelle voorgestel wat gebruik kan word om die onderskeie alternatiewe te evalueer en die beste opsie te selekteer. Die eerste model evalueer die kostes wat geassosieër word met elk van die verskillende vervoeropsies. Die tweede model stel die gebruiker instaat om die kostes van verskillende Verspreidingsentrums te vergelyk. Die derde model is ‘n multi-kriteria besluitnemingsmodel wat HM Timber kan help om die beste opsie vir elke meul en elke streek te selekteer. Die VISA vervoermodel dui aan dat Manline die beste opsie is vir alle aflewerings vanaf Weza en Singisi. Manline word ook aangedui as die beste opsie vir Langeni se langafstand aflewerings, terwyl HM Timber se eie vloot steeds die beste opsie is vir Langeni se plaaslike aflewerings. Resultate van die Verspreidingsentrum-model dui aan dat die kostestrukture baie sensitief is vir veranderinge in die premies wat behaal word asook veranderinge in die houtmark, insluitend veranderinge in die kliëntebasis, die tipe produkte en die volumes wat verkoop word. Die model toon aanvaarbare winsgrense vir Verspreidingsentrums in Johannesburg en Port Elizabeth, maar die geprojekteerde winsgrens vir Durban is minder as 2 %, wat beteken dat ‘n Verspreidingsentrum in Durban nie ‘n goeie opsie is nie. ‘n Direkte verspreidingsoplossing word eerder vir die KwaZulu-Natal streek voorgestel. Hierdie opsie behels meervoudige aflewerings wat direk vanaf die verskillende meulens gemaak sal word. Agt faktore word uiteindelik geïdentifiseer as noodsaaklik vir die suksesvolle vervoer en verspreiding van houtprodukte. HM Timber vaar baie goed met die meeste van hierdie faktore, maar enkele punte moet aangespreek moet word ten einde huidige diensvlakke te handhaaf. Hierdie punte sluit in investering in ‘n toegewyde logistieke afdeling, verbetering van interne en eksterne samewerking en die verbetering van ‘n aantal operasionele- en inligtingstelsels.

Dissertation (PhD)--Stellenbosch, 2004. / ENGLISH ABSTRACT: The dissertation reflects the research done on the design principles and success factors for the operation of cross dock facilities in grocery and retail supply chains. The cross dock is a particular facility in the supply chain where goods are received from suppliers, sorted without storage of the goods, and then efficiently moved to downstream customers. Cross docks are not a new operation. However, the use in high volume grocery and retail operational capabilities is poorly understood and is not uniquely defined. The problem is that cross docks are often seen as extensions of warehouses. The same personnel, systems and processes are applied and the efficiency potential of the cross dock is not achieved. Warehouses are orientated towards storing the full range of product and allowing the pick to be done from this storage buffer to provide any or all of these products to a customer. Cross docks will only handle products that are used in larger quantities and that are sent to most, if not all, the customers. The cross dock is therefore distinct and very different from the traditional warehouse. The published research tends to focus on the technical aspects of the cross dock layout. This research is primarily in the scheduling of the trucks into the yard of the facility; the allocation of trucks to specific doors of the facility; and the allocation of doors to receiving and despatch functions within the facility. Very little information or research reflects the design principles and success factors for the cross dock and its supply chain. The only classification of the cross dock in the literature is whether the barcode is added to the item before or after receipt at the cross dock. For this research work a literature survey was conducted and five major operations were reviewed, in South Africa and the USA. The research empirically drew logical conclusions, which were tested in the operations and found to be correct. This allowed the design principles and success factors to be determined for a successful cross dock. The research extends the knowledge of the cross dock operation and design: - • A new classification for the feasible types of cross docks in the supply chain was developed. Three factors are shown to be of primary importance: - o Where in the supply chain the identification of specific items for a customer is done; o Where the sort is done for the items to be delivered to a customer; and o Whether the supplier is providing one product or multiple products to the sort. From these three factors, eight potential classifications could be defined. However, only three practical types of cross dock can be determined from these eight alternatives. These are named in this research as Cross Dock Managed Load (CML); Joint Managed Load (JML); and the Supplier Managed Load (SML). The cross dock is far more effective than the warehouse when the total work (excluding inventory) is considered. The earlier in the supply chain the product is identified for the use of the entire downstream supply chain, the more effective will be the total supply chain. Thus the greatest supply chain effectiveness possible is with the SML, then the JML and finally the CML. • The operation of a cross dock is very similar to a continuous manufacturing process. There is no buffer of stock to decouple the inbound and outbound processes, and the operation takes place in a restricted area. However, in the retail chain, the workload alters with different orders and different days. Daily load differences vary by as much as 90%. This results in vastly different workloads and variations of throughput. This is similar to a batch operation with highly variable workloads between batches. The literature recommends the use of Just in Time (JIT) practice for cross docks. This is inappropriate as its primary requirements are continuous full volume operation and continuous small improvements to achieve a balanced operation. The most appropriate method of process improvement is the Theory of Constraints (TO C) and not JlT. • The management must have a detailed, disciplined approach. This implies standardised methods of operation, and a high degree of training. Equally there is the requirement for a special type of personnel to operate the cross dock. These operating personnel must be able to operate with precision (i.e. very low error rates) and be able to maintain this capability for continuous periods. • The systems required for a successful operation must include the capabilities of Yard Management, WMS for cross docking, Order Management with Advanced Shipping Notice (ASN) capability and Track and Trace across the supply chain. The items need to be identified by a barcode. The information required on the barcode will be determined by. the information systems capability of the least advanced service provider in the supply chain. If this service provider can receive and transmit all the data required for the supply chain from and to the other members, then the barcode need only be an identification number of the specific item. The data pertaining to the items is then passed from system to system in the supply chain. If data movement is not possible between all the parties in the entire supply chain, then the barcode must contain the information that will identify the item, the origin and the final delivery destination. If the items are delivered as part of a consignment, a further quantum of information is required to identify the total number of items in the consignment and the specific item within the consignment. • The research shows that the overall capability of the cross dock or its maximum capacity is the combination of the capability of the personnel and the cross dock design. Restrictions on either the personnel capability or the design of the cross dock, or both, severely reduces the effectiveness of the cross dock. • The previous research on the sequence of allocation of trucks to specific doors within the cross dock can be enhanced with a new sequencing method. The new method allocates the transport, in sequence of arrival, to the open door that either numrruses the walk distance in the facility; or maximises the completion of the consignments in order to minimise the area required to build the consignments; or a combination of both. The choice of these will be determined by the constraints imposed by the design of the building. This is an important extension as this ties the supply chain into the cross dock operation, rather than looking at the cross dock in isolation as has been done in this previous research. • The factors that influence the design of a cross dock as to its size, shape, number of doors, and the specifically required additional areas, is defined in detail. The principles of these factors and their inter-relationships and dependencies are used in a detailed design for a cross dock. The detailed design process is set out from data analysis through to the actual size calculations and layouts. Measurements of walk distance and sort movement are used to determine the most effective design. The design is shown to be considerably more effective than the older designs. This work has significantly extended the research on the design principles and success factors for implementation of cross docks in retail supply chains. The research derives a unique new classification for cross docks. An improvement is made to existing research on the allocation of the transport to particular doors in the cross dock. The operation, management and personnel are shown to require specific characteristics. The information systems required for effective cross docks is determined and defined. The identification of the individual items by barcode and the information required within the barcode depending on the information sophistication of the service providers in the supply chain is defined. A detail process to design a cross dock is evolved, with the full knowledge of the factors that must be considered and their interrelationships. Measurements to determine the effectiveness of the design are used to choose the most appropriate design. All these are then synthesised into a new design, which is far more effective than any of the other designs researched. The design process will produce a very effective cross dock as has been demonstrated with a new facility. / AFRIKAANSE OPSOMMING: geen opsomming

published_or_final_version / abstract / toc / Industrial and Manufacturing Systems Engineering / Master / Master of Philosophy

D. Tech. (Business, Logistics, Faculty of Management Sciences), Vaal University of Technology / The South Africa economy has embraced the importance of small and medium enterprises (SMEs) as agents of economic growth and sustainability. However, for SMEs to be more efficient and effective towards the growth of the economy, it is important for SMEs to implement and take cognisance of the global competitive strategy among the so-called “best in class” organisations in order to also sustain and grow their business. The supply chain network with its flexibility has become the most commonly used business strategy worldwide for the improvement of performance in organisations, more especially among larger organisations. In today’s competitive environment, successful organisations are those that have been able to link their business functions within the organisation itself as well as across other businesses outside the organisation. Competitive challenges in the modern business environment have resulted in the need for organisation to integrate business processes strategically across other business units within the supply chain network. Network perspective theory, social network theory, network management theory and relational view theory are discussed to better understand the importance of SMEs supply chain network, flexibility and integration. The purpose of this study was to examine the influence of supply chain network, flexibility and integration on the SMEs business performance in the Southern Gauteng region. The research survey was conducted in the Meyerton, Vereeniging and Vanderbijlpark and 401 SMEs participated in the study. SPSS 22.0 was used to analyse the data and AMOS 22.0 was used to perform the confirmatory factor analysis. The structural equation modelling (SEM) was used to assess the proposed model fit and to test the statistically significant relationship of the hypothesis. The research study results revealed that supply chain network, flexibility and integration positively influence SMEs business performance. This study contributes new knowledge to the existing literature by providing a research framework that can enhance SMEs performance and also provide practical recommendations based on the research findings for SMEs and for future research. Furthermore, as one of the first studies addressing the influence of supply chain network, flexibility and integration on the performance of SMEs in the southern Gauteng region it has generated new insights and information as well as outlined the strategic reasons for SME owners and managers to improve on their business relationships. / VUT

D.Com. (Transport Economics) / Please refer to full text to view abstract

Esta dissertação trata do planejamento do abastecimento de última milha em centros urbanos, propondo métodos para agrupar pedidos de clientes em programação de entregas. Neste estudo, é considerado que o frete pago ao transportador em uma rota é definido pela distância direta do ponto de entrega mais distante do depósito em contraposição à distância total da rota que é usual na literatura sobre problemas de roteirização de veículos. Além disso, também são consideradas categorias, conjunto de produtos similares, que não podem ser transportadas juntas por não serem compatíveis entre si. O objetivo do problema proposto é determinar o agrupamento e sequenciamento de pedidos em roteiros de veículos de acordo com as características operacionais descritas acima, utilizando uma frota homogênea de veículos capacitados que parte de um depósito, de tal forma que toda a demanda seja atendida com o menor frete possível. Para resolução desse problema são propostas uma formulação matemática para obtenção de soluções exatas e a implementação da heurística \"Multi Start Perturbation Tabu\" (MSPT) que é composta das metaheurísticas \"Greedy Randomized Adaptive Search Procedure\" (GRASP), \"Tabu Search\" (TS) e \"Iterated Local Search\" (ILS) para obtenção de soluções heurísticas. Os resultados experimentais indicam que a MSPT é competitiva com os resultados do método exato com até 5 horas de processamento utilizando os recursos computacionais de alto desempenho do Laboratório de Computação Científica Avançada (LCCA) da Universidade de São Paulo. / This dissertation addresses the planning of the last mile supply in urban centers and proposes methods to group customer orders into shipments. In this study, freight paid to the carrier on a route is defined as the direct distance from the point of delivery that is furthest from the depot as opposed to be defined as the total distance of the route which is commonly found in the literature on vehicle routing problems. In addition, it is also considered categories, a set of similar products, which cannot be transported together because they are not compatible with each other. The objective of the proposed problem is to determine the grouping and sequencing of orders into vehicle shipments according to the operational characteristics described above, using a homogeneous fleet of capacitated vehicles that is located in a depot, in such a way that all the demand is delivered with the lowest freight possible. To solve this problem, it is proposed a mathematical formulation to obtain exact solutions and the implementation of the Multi Start Perturbation Tabu (MSPT) heuristic that is composed of the Greedy Randomized Adaptive Search Procedure (GRASP), Tabu Search (TS) and \"Iterated Local Search\" (ILS) for heuristic solutions. Finally, the experimental results indicate that the MSPT is competitive with the outcomes of the exact method with up to 5 hours of processing using the high performance computational resources of the Advanced Scientific Computation Laboratory (LCCA) of the University of São Paulo (USP).

by Yang Ming-Yen Teresa. / Thesis (M.B.A.)--Chinese University of Hong Kong, 1991. / Includes bibliographical references (leave 61). / ABSTRACT --- p.ii / ACKNOWLEDGEMENTS --- p.iii / TABLE OF CONTENTS --- p.iv / LIST OF ILLUSTRATIONS --- p.vi / LIST OF TABLES --- p.vii / Chapter I. --- Introduction --- p.1 / Company Background --- p.6 / Chapter II. --- Methodology --- p.9 / Chapter III. --- Analysis of the Existing System --- p.10 / Piecegoods Purchase Order System --- p.10 / Physical Flow --- p.17 / Yarn Dyeing --- p.24 / Knitting --- p.26 / Piece Dyeing --- p.26 / Finishing --- p.30 / Final Inspection --- p.32 / Transit Time --- p.35 / Chapter IV. --- Identification of Major Problem Areas --- p.38 / Documentation Flow --- p.38 / Physical Flow --- p.40 / Feasibility of the Target Lead Time --- p.41 / Feasibility of the Target --- p.42 / Production Capacity / Identification of the Problem Areas --- p.44 / Chapter V. --- Recommendations To Improve the Lead Time --- p.50 / Measurements --- p.50 / Skills --- p.51 / Communication --- p.54 / Decision Process --- p.56 / Management Support --- p.57 / Appendix --- p.60 / Bibliography --- p.61

To Chi Kit. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2006. / Includes bibliographical references (leaves 50-52). / Abstracts in English and Chinese. / Abstract --- p.i / Acknowledgement --- p.iii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Structure of thesis --- p.3 / Chapter 2 --- Literature Review --- p.4 / Chapter 3 --- Problem description and formulation --- p.7 / Chapter 3.1 --- Notation --- p.9 / Chapter 3.2 --- Cost Structure --- p.11 / Chapter 3.3 --- Assumptions --- p.12 / Chapter 3.4 --- Problem Formulation --- p.14 / Chapter 4 --- Stations with deterministic demand --- p.15 / Chapter 4.1 --- Greedy Algorithm --- p.15 / Chapter 4.2 --- Example --- p.16 / Chapter 4.3 --- Properties --- p.17 / Chapter 5 --- Stations with stochastic demand --- p.21 / Chapter 5.1 --- Decision planned before arrival --- p.26 / Chapter 5.2 --- Decision made after vehicle arrival --- p.29 / Chapter 6 --- Numerical example --- p.38 / Chapter 6.1 --- Comparing decision made before and after arrival of sta- tion --- p.39 / Chapter 6.2 --- Relation between K and li --- p.40 / Chapter 6.3 --- Relation between unit penalty / cost value with K . . . --- p.40 / Chapter 7 --- Conclusion --- p.47