Decision support for fresh produce replenishment order schedules in a retail outlet

Lotter, Janneke (2017-03)

Thesis (MEng)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: Retailers are faced with complex periodic decisions related to replenishment orders for fresh produce and the allocation of shelf space to these products. The complexity of these decisions may be attributed to the short shelf lives of fresh produce, a variety of operational constraints and a limitation on the amount of shelf space available. Since retailers may gain a competitive advantage through the e cient management of shelf space in their fresh produce departments, an objective methodology is required to provide decision support for fresh produce replenishment order decisions. This is especially important for retail outlets with little or no backroom storage space for their fresh produce, where all product units that arrive at the store have to be displayed on the shelves immediately. A mathematical model is proposed in this thesis in support of the aforementioned fresh produce replenishment order decisions of retailers. The aim of the model is to satisfy forecasted product demand over a prescribed decision period as closely as possible, while maximising pro t and minimising product waste | all subject to a xed amount of available shelf space. Two solution approaches are designed for solving this model which takes the form of a mixed integer programming problem. An exact solution approach is implemented in CPLEX for solving the model in the context of small, hypothetical problem instances. Due to the computational complexity of the exact solution approach, a (metaheuristic) solution approach, based on the method of simulated annealing, is also established by which larger, realistic instances of the model can be solved much quicker, albeit approximately. The mathematical model and the approximate solution approach are embedded in a computerised decision support system concept demonstrator so as to provide managers of retail outlets with a practical tool able to recommend fresh produce replenishment order schedules in line with the aforementioned retail objectives. A real-life case study is performed, involving 72 products in the ambient section of the fresh produce department within a retail outlet in Grassy Park, Cape Town, so as to demonstrate the practicality of the decision support system and illustrate the high quality of its recommendations. Upon comparing the replenishment order recommendation of the system with that actually employed at the outlet, it is found that the system yields substantially superior results.

AFRIKAANSE OPSOMMING: Kleinhandelaars staar moeilike besluite rakende aanvullingsbestellings van varsprodukte en die toewysing van rakspasie aan hierdie produkte in die gesig. Die kompleksiteit van hierdie besluite kan toegeskryf word aan die kort rakleeftye van hierdie produkte, verskeie operasionele beperkings en beperkte beskikbare rakspasie. Aangesien kleinhandelaars 'n kompeterende voordeel mag trek uit die doeltreffende bestuur van rakspasie in hul varsprodukafdelings, word 'n objektiewe metodologie vereis waarvolgens hulle met besluitsteun in terme van varsprodukaanvullingsbesluite bedien kan word. Hierdie vereiste is veral belangrik in gevalle waar kleinhandelwinkels beperkte of geen stoorkamerspasie vir varsprodukte tot hul beskikking het nie, en waar al hierdie produkte dus dadelik op die rakke ten toon gestel moet word. 'n Wiskundige model word in hierdie tesis daargestel waarvolgens steun in terme van die bogenoemde varsprodukaanvullingsbesluite aan kleinhandelaars gebied kan word. Die doel van die model is om sover moontlik aan vooruitgeskatte aanvraag vir hierdie produkte oor 'n voorgeskrewe beslissingsperiode te voldoen, terwyl wins gemaksimeer word en produkvermorsing geminimeer word | alles onderhewing aan 'n vaste hoeveelheid beskikbare rakspasie. Twee oplossingsbenaderings word ontwerp waarvolgens hierdie model, wat die vorm van 'n gemengde heeltallige programmeringsprobleem aanneem, opgelos kan word. 'n Eksakte oplossingstegniek word in CPLEX geïmplementeer waarmee klein, hipotetiese probleemgevalle opgelos kan word. As gevolg van die berekeningskompleksiteit van die eksakte oplossingsbenadering word 'n (metaheuristiese) oplossingstegniek gebaseer op die metode van gesimuleerde tempering ook daargetsel waarmee groter, realistiese gevalle van die probleem vinniger, maar benaderd, opgelos kan word. Die wiskundige model en die benaderde oplossingstegniek word in 'n gerekenariseerde besluitsteun-konsepdemonstrator ingesluit om sodoende kleinhandelaars van 'n praktiese hulpmiddel te voorsien wat daartoe in staat is om aanbevelings in terme van varsprodukaanvullingsbesluite volgens die bogenoemde kleinhandeldoelwitte te maak. 'n Werklike gevallestudie word uitgevoer waarin 72 produkte wat in die kamertemperatuur-gedeelte van die varsprodukafdeling van 'n kleinhandelwinkel in Grassy Park, Kaapstad in ag geneem word om sodoende die praktiese nut van die besluitsteunstelsel en die hoë kwaliteit van die aanbevelings wat die stelsel genereer, te illustreer. Deur die aanvullingskedule wat die stelsel aanbeveel met diè wat werklik by die winkel gevolg is, te vergelyk, word daar bevind dat die stelsel noemenswaardige beter resultate lewer.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/101082
This item appears in the following collections: