Multi-objective optimisation in the retail banking industry with stochastic discrete-event simulation

Scholtz, Esmarie (2011-12)

How can the service level of the researched ATM network, operated by the particular bank, be maximised at minimum cost using industrial engineering techniques?

Final year project (BEng)--Stellenbosch University, 2011.

Technical Report

ENGLISH ABSTRACT: Cash management is a multi-objective optimisation problem which aims to maximise the service level provided to customers at minimum cost. The topic of du Toit's (2011) masters thesis was automated teller machice (ATM) cash management for a specific South African retail bank. Focus was placed on an ATM network which primarily provides cash to blue collar laborers in the rural Eastern Cape. The aim of this final year project is to refine the work done by du Toit through the specific investigation into the effect of applying a combination of the vehicle routing problem (VRP) and continuous review policy for inventory management to the retail banking industry. A decision support system (DSS) in the form of a stochastic, discrete-event simulation model is developed. 90 different scenarios are experimented with using the DSS. Results show that the application of the VRP consistently yields high service levels at low cost when compared to two other routing approaches: first-in- first-out routing and direct replenishment. It is concluded that use of the VRP is especially beneficial when the bank has substantial control over transportation cost. The principal recommendation is therefore that cost control should be maximised to fully exploit the advantages obtainable from effective cash management. Finally, it is argued that the benefits to be gained from effective cash management (higher service levels at lower cost) can lead to the improvement of the lives of many a South African wage earner. These benefits could also lead to an increased profit margin - life is all about choices.

AFRIKAANSE OPSOMMING: Kontantbestuur is multi-doelstelling optimeringsprobleem waarvan die doelwit is om die diensvlak wat aan kliënte gelewer word, te maksimeer, terwyl koste minimeer word. Die onderwerp van du Toit (2011) se meesters tesis was outomatiese tellermasjien (OTM) kontantbestuur vir 'n spesifieke Suid-Afrikaanse kleinhandelbank. Fokus is geplaas op 'n OTM netwerk wat hoofsaaklik kontant aan arbeiders in die landelike Oos-Kaap voorsien. Die doel van hierdie finale jaar projek is om du Toit se werk te verfyn deur spesifiek ondersoek in te stel na die effek wat die toepassing wat 'n kombinasie van die voertuigskeduleringsprobleem en die deurlopende hersieningsbeleid vir voorraadbestuur sal hê. 'n Stogastiese, diskrete-gebeurtenis simulasie model is ontwikkel om as besluitnemings ondersteuningstelsel te dien. 90 verskillende eksperimente is met die simulasie model voltooi. Resultate toon dat die toepassing van die voertuigskeduleringsprobleem deurlopend hoë diensvlakke teen vergelykende lae koste tot gevolg het. Die voertuigskeduleringsprobleem word vergelyk met twee ander skeduleringstegnieke: eerste-in-eerste-uit skedulering en direkte aanvulling. Die gevolgtrekking word gemaak dat gebruik van die voertuigskeduleringsprobleem veral voordelig is wanneer die bank aansienlike beheer oor vervoerkoste het. Die hoof aanbeveling is daarom dat kostebeheer gemaksimeer behoort te word om ten volle munt te slaan uit die voordele wat moontlik gemaak word deur effektiewe kontantbestuur. Ten slotte word daar aangevoer dat die voordele wat sal volg uit effektiewe kontantbestuur (hoër diensvlakke teen laer koste), die lewens van vele Suid-Afrikaanse loonwerkers kan verbeter. Die voordele kan ook lei tot 'n vergrote winsmarge -- die lewe is vol keuses.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/18137
This item appears in the following collections: