In order to solve the customer service delay and financial risk growth problems due to transaction delays in the integrated payments system (IPS) at a large Canadian Bank, a stochastic simulation model was built using the GPSS/PC simulation language and custom FORTRAN subroutines. These tools were used to analyze the present system and critical parameters, such as service rates of the operators at each suboperation, transaction arrival rates, and arrival patterns were identified and estimated where necessary. Three alternatives, including two automation strategies to increase processing efficiency and a control system to ensure a guaranteed service time were developed which, if implemented, will improve the operation of the IFS. Analysis of the simulation output showed that no single alternative dominates the others although each alternative, on its own, improved the present system with respect to the performance measures studied. A detailed examination of the costs and benefits of each alternative should be completed before the Bank decides on the appropriate strategy for final implementation.
[1]
D. Fakinos.
The Infinite Server Queue with Arrivals Generated by a Non-Homogeneous Compound Poisson Process
,
1984
.
[2]
Emanuel Parzen,et al.
Stochastic Processes
,
1962
.
[3]
James Essinger.
Managing technology in financial institutions
,
1993
.
[4]
Bernard O. Koopman,et al.
Air-Terminal Queues under Time-Dependent Conditions
,
1972,
Oper. Res..
[5]
P. A. Bobillier,et al.
Simulation With Gpss and Gpss V
,
1976
.
[6]
Peter A. W. Lewis,et al.
Statistical Analysis of Non-Stationary Series of Events in a Data Base System
,
1976,
IBM J. Res. Dev..
[7]
Edwin G. Landauer,et al.
Reducing Waiting Time at Security Checkpoints
,
1989
.
[8]
Peter Kolesar,et al.
Stalking the Endangered CAT: A Queueing Analysis of Congestion at Automatic Teller Machines
,
1984
.