This document discusses optimizing the queueing system of an HDFC Bank ATM using queueing theory. It analyzes the arrival and service rates on weekdays during busy and free schedules as well as weekends to calculate key queueing parameters like utilization factor, average queue length, and waiting time. The M/M/1 queueing model is used to solve the problem and find an optimal solution. Calculations show the ATM is fully occupied during prime hours with a utilization factor close to 1. During non-busy hours and weekends, the utilization is lower at 60% and 70% respectively. This analysis can help determine if one or multiple ATM machines are needed to minimize customer wait times.
Download as DOC, PDF, TXT or read online on Scribd
0 ratings0% found this document useful (0 votes)
65 views
Or Report
This document discusses optimizing the queueing system of an HDFC Bank ATM using queueing theory. It analyzes the arrival and service rates on weekdays during busy and free schedules as well as weekends to calculate key queueing parameters like utilization factor, average queue length, and waiting time. The M/M/1 queueing model is used to solve the problem and find an optimal solution. Calculations show the ATM is fully occupied during prime hours with a utilization factor close to 1. During non-busy hours and weekends, the utilization is lower at 60% and 70% respectively. This analysis can help determine if one or multiple ATM machines are needed to minimize customer wait times.
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 11
Optimization of HDFC Bank ATM using Queueing Theory
A report submitted in partial fulfillment of the
reuirements for the !ourse of O"#$AT%O&' $#'#A$CH
by &AM#( AA)*'H A+A$,A- $#+. &O( //BM#0012 'CHOO- OF M#CHA&%CA- A&D B*%-D%&+ 'C%#&C#' MA) 30/1 1 CO&T#&T' Chapter No. Title Page No. CHA"T#$ / %&T$OD*CT%O& 4 /./ Ob5e!ti6es 7 ,hole !ontent of the pro5e!t /.3 %ntrodu!tion to the industry 1 CHA"T#$ 3 -%T#$AT*$# $#8%#, 9 CHA"T#$ 4 M#THODO-O+) : 4./ "roblem 'tatements 4.3 'olutions 2 CHA"T#$ 1 $#'*-T' A&D D%'C*''%O& ; 1./ $esults and its signifi!an!e 1.3 Ad6antages < Appli!ations CHA"T#$ 9 CO&C-*'%O&' ; $#F#$#&C#' /0 2 /. %&T$OD*CT%O& Queue is a common word that means a waiting line or the act of joining a line. Queuing theory was initially proposed by A.. !rlang in 1"#$. %t optimi&es the number of ser'ice facilities and adjusts the times of ser'ices. Queuing theory is the study of (ueue or waiting lines. )ome of the analysis that can be deri'ed using (ueuing theory include the e*pected waiting time in the (ueue+ the a'erage time in the system+ the e*pected (ueue length as well as the probability of the system to be in certain states+ such as empty or full. Queuing is the common acti'ity of customers or people to a'ail the desired ser'ice+ which could be processed or distributed one at a time. AT, is an automatic teller machine which is used to sa'e the cost and reachability of a ban- by satisfying customer needs. Customers can withdraw and deposit money without any paperwor- and it facilitates them to reduce time and cost to go to ban- in person. .an- AT,s would a'oid losing their customers due to a long wait on the line /./ OB=#CT%8# A&D ,HO-# CO&T#&T OF TH# "$O=#CT %n this project + (ueuing theory is applied to a /01C .an- AT, in Pune. )ome of the analysis that can be deri'ed using (ueuing theory include the e*pected waiting time in the (ueue+ the a'erage time in the system+ the e*pected (ueue length as well as the probability of the system to be in certain states+ such as empty or full ,odel % 2,3,314 is used for sol'ing+ this model deals with a (ueueing system ha'ing Poisson input+ e*ponential ser'ice and there is no limit on the system capacity while the customers are ser'ed on a first come +first basis. "roblem ob5e!ti6e< The o'erall objecti'e of this project is to reduce the waiting time of customers by pro'iding an optimum solution to the problem. Weekdays Busy Schedule HDFC (time in sec) $ Weekdays Free Schedule HDFC (time in sec) Weekends Schedule HDFC (time in sec) /.3 %&T$OD*CT%O& TO %&D*'T$) 0ata is obtained from the branch manager+ /01C .an-+ /injewadi .ranch+ Pune+ ,aharashtra. /01C .an- 5imited is an %ndian financial ser'ices company based in ,umbai+ ,aharashtra. %t was incorporated in 1""6.
/01C .an- is the fifth largest ban- in %ndia by assets. %t is the largest ban- in %ndia by mar-et capitali&ation as of 26 1ebruary 2#16. 6 3. -%T#$AT*$# $#8%#, 7asumathi.A+ 0hana'anthan P 22#1#4+ in their research paper on Application of )imulation Techni(ue in Queuing ,odel for AT, 1acility+ the authors formulated a suitable simulation techni(ue which will reduce idle time of ser'ers and waiting time of customers for any ban- ha'ing AT, facility. This research paper would be helpful for any ban- at global le'el for impro'ing their customer ser'ice towards competition. A&mat Nafees+ Q8!8%N9 T/!:;< AN0 %T) APP5%CAT%:N=ANA5<)%) :1 T/! )A5!) C/!C:8T :P!;AT%:N %N %CA )8P!;,A;!T+ 22##>4+ This paper contains the analysis of Queuing systems for the empirical data of supermar-et chec-out ser'ice unit as an e*ample. :ne of the e*pected gains from studying (ueuing systems is to re'iew the efficiency of the models in terms of utili&ation and waiting length+ hence increasing the number of (ueues so customers will not ha'e to wait longer when ser'ers are too busy. ;yan .erry+ Queueing Theory+ this paper defines the building bloc-s and deri'es basic (ueueing systems. %t begins with a re'iew of some probability theory and then defines processes used to analy&e (ueuing systems+ in particular the birth?death process. A few simple (ueues are analy&ed in terms of steady?state deri'ation before the paper discusses some attempted field research on the topic ). . 0har1+ Tan&ina ;ahman 22#1$4+ in their research paper on Case study for .an- AT, Queueing model+ calculated parameters li-e utili&ation factor+ (ueue length and waiting time. .ased on this calculation a suitable conclusion is drawn and the result can help in better functioning of the ban-. .ha'in Patel+ Pra'in .hathawala 3 %nternational @ournal of !ngineering ;esearch and Applications 2%@!;A4 %))N= 226A?"B22+ 7ol. 2+ %ssue C+ )eptember? :ctober 2#12+ pp.12>A? 12A6. Case )tudy for .an- AT, Queuing ,odel. ,athias0harmawirya+ !rwin Adi+ %nternational Conference on ,anagement and Artificial %ntelligence %P!0; 'ol.B 22#114 D 22#114 %AC)%T Press+ .ali+ %ndonesia. The paper discusses the benefits of applying (ueueing analysis to a busy restaurant. Crispin !mmanuel 0. 0ia&+ @im @oel ,adrigal+ Proceedings of the !astern Asia )ociety for Transportation )tudies+ 7ol. C+ pp. 16"B ? 1C#"+ 2##C+ A55:CAT%:N :1 !5!CT;:N%C T:55 C:55!CT%:N 5AN!) AT T:55 P5AEA) C:N)%0!;%N9 ):C%A5 :PT%,%EAT%:N :1 )!;7%C! T%,!) AN0 0!5A<+ %n this paper (ueueing analysis was done and an optimum method was implemented to reduce the waiting period. 9er oole F A'ishai ,andelbaum+ Queueing ,odels of Call Centers 22##14+ This is a sur'ey of some academic research on telephone call centres. Queueing model is used for efficient call centre management. C 4. M#THODO-O+) 2,3,3%4 (ueueing model is used to sol'e the problem and find an optimal solution. ,3,31 (ueue represents the (ueue length in a system ha'ing a single ser'er+ where arri'als are determined by a Poisson process and job ser'ice times ha'e an e*ponential distribution. An ,3,31 (ueue is a stochastic process whose state space is the set G#+ 1+ 2+ $+...H where the 'alue corresponds to the number of customers in the system+ including any currently in ser'ice. Arri'als occur at rate I according to a Poisson process and mo'e the process from state i to i J 1. )er'ice times ha'e an e*ponential distribution with parameter K in the ,3,31 (ueue. A single ser'er ser'es customers one at a time from the front of the (ueue+ according to a first?come+ first?ser'ed discipline. Lhen the ser'ice is complete the customer lea'es the (ueue and the number of customers in the system reduces by one. The buffer is of infinite si&e+ so there is no limit on the number of customers it can contain. 4./ "roblem 'tatement %n most of the AT,s the major problem is waiting of customers in the (ueue for more duration. ,ainly the objecti'e of AT, for ban- is to -eep away the customers from coming to ban- and ma-e the process easy for them to a'oid the basic procedure they do in ban-. .ut the problem which most AT, face is the long (ueue in front+ but then when the problem is only for a short while as rest of the time the AT, remains idle means adding to the operating cost. The problem is to determine whether only one machine is re(uired to fulfill the need or two more machines needed to be installed to gi'e comfort to customer which is really of short period of time. B > 4.4 'olutions 1rom the abo'e obtained data+ Mean Arrival time 1 customer in 47 sec () weekdays busy Mean Servicetime 1 customer in 95 sec () weekdays busy weekdays ree 1 customer in !1" sec #eekdays ree 1 customer in 1!" sec weekends 1 customer in 166 sec #eekends 1 customer in 9$ sec 1) =/ (wee-days busy4 =0.6315/1.299 =0.486 =0.285/0.5=0.57 2wee-days free4 =0.4167/0.625=0.66672 2wee-ends4 2) P # M1? ( wee-days busy4 =10.486 =0.514 P # M1?#.C> 2 wee-days free4 M#.6$ P # M1?#.BBB>2 2wee-ends4 M#.$$$2A
$4 5M 1.26BB32#.B$1C?1.26BB4 M 2.#2B (wee-days busy4
5M #.2AC32#.C?#.2AC4 M 1.$2CB 2wee-days free4
5M #.61B>32#.B2C?#.61B>4 M 1.""2A 2wee-ends4
64 5(M #.6AB*2.#2B M #."A6 (wee-days busy4 A 5(M#.C>*1.$2CB M #.>CCC 2wee-days free4 5(M#.BBB>2*1.""2A M 1.$2A 2wee-ends4 C4 L(M#."A631.26BB M #.>A" mins (wee-days busy4 L(M#.>CCC3#.2AC M 2.BC#A mins 2wee-days free4 L(M 1.$2A3#.61B> M $.1B" mins 2wee-ends4 B4 LM 2.#2B31.26BB M 1.B2C mins (wee-days busy4 LM 1.$2CB3#.2AC M 6.BC122 mins 2wee-days free4 LM 1.""2A3#.61B> M 6.>A2$ mins 2wee-ends4 1. $#'*-T' A&D D%'C*''%O&' 7arious parameters li-e a'erage (ueue length+ waiting time+ utili&ation factor is calculated. I is calculated by considering arri'al time of the customers to that of the number of customers. )er'ice time is the time re(uired for completion of a ser'ice i.e.+ it is the time inter'al between beginning of a ser'ice from AT, machine and its completion. .ased on the calculation during prime hours AT, is fully occupied+ thus implied by utili&ation factor appro*imately e(ual to 1. 0uring non?busy hours it is 6$N and wee-ends it is $$N. :ne of the major problem is O:ut of ser'iceP. Assumption made for calculation is that the customers donQt lea'e the (ueue before the ser'ice. 1./ Ad6antages This case study can help ban- AT, to increase its Qo) 2Quality of )er'ice4+ by anticipating+ if there are many customers in the (ueue. The result of this case study may become the reference to analy&e the current system and impro'e the ne*t system. .ecause the ban- can now estimate the number of customers waits in the (ueue and the number of customers going away each day. .y estimating the number of customers coming and going in a day+ the ban- can set a target that+ how many AT,s are re(uired to ser'e people in the main branch or any other branch of the ban-. 9.CO&C-*'%O& The main purpose of this study is to de6elop an effi!ient pro!edure for ATM ueuing problem> ?hi!h !an be daily used by banks to redu!e the ?aiting time of !ustomers in the system. 8arious parameters of ueueing are !al!ulated. %t is !on!luded from the results that in order to redu!e the ?aiting time a ne? ATM should be installed. This ?ill be helpful for the bank to sustain more potential !ustomers in !ompetiti6e situations ?ith other banks. " $#F#$#&C#' @/1 Bold CapsA @ To be arranged in year?iseA /. 7asumathi.A 1 + 0hana'anthan P 2
1 Assistant Professor+ 7%T .usiness )chool+ 7%T 8ni'ersity+ Chennai. 2 Professor and /ead+ 0epartment of )tatistics+ Pondicherry 8ni'ersity+ Pondicherry Application of )imulation Techni(ue in Queuing ,odel for AT, 1acility+ %NT!;NAT%:NA5 @:8;NA5 :1 APP5%!0 !N9%N!!;%N9 ;!)!A;C/+ 0%N0%985+ 7olume 1+ No $+ 2#1# 3. ). . 0har1+ Tan&ina ;ahman2 Case )tudy for .an- AT, Queuing ,odel+ %:); @ournal of ,athematics 2%:);?@,4 e?%))N= 22>A?C>2A+p?%))N= 2$1"?>BCR+ 7olume >+ %ssue 1 2,ay. S @un. 2#1$4+ PP #1?#C 4. 9er oole1 F A'ishai ,andelbaum2 17rije 8ni'ersiteit+ 0e .oelelaan 1#A1a+ 1#A1 /7 Amsterdam+ The Netherlands 2%ndustrial !ngineering and ,anagement+ Technion+ /aifa $2###+ %srael+ Queueing ,odels of Call Centres+ An %ntroduction. 6. Crispin !mmanuel 0. 0ia&+ @im @oel ,adrigal+ Proceedings of the !astern Asia )ociety for Transportation )tudies+ 7ol. C+ pp. 16"B ? 1C#"+ 2##C+ A55:CAT%:N :1 !5!CT;:N%C T:55 C:55!CT%:N 5AN!) AT T:55 P5AEA) C:N)%0!;%N9 ):C%A5 :PT%,%EAT%:N :1 )!;7%C! T%,!) AN0 0!5A< C. ,athias0harmawirya+ !rwin Adi+ %nternational Conference on ,anagement and Artificial %ntelligence %P!0; 'ol.B 22#114 D 22#114 %AC)%T Press+ .ali+ %ndonesia. B. A&mat Nafees+ Q8!8%N9 T/!:;< AN0 %T) APP5%CAT%:N=ANA5<)%) :1 T/! )A5!) C/!C:8T :P!;AT%:N %N %CA )8P!;,A;!T >. 9er oole F A'ishai ,andelbaum+ Queueing ,odels of Call Centers 22##14 11