A Queue Fairness Model of Automated Teller Machine ATM System A Case Study of Financial Institutions in Kaduna Nigeria
Until recently, consensus of opinions, thoughts and theories on the studies of ATM queuing system, and the analysis of its influence on customers’ behaviour were predicated on the average time spent by a tagged customer on the queue. Very few were poised to consider the unfairness or injustice experienced by customers in the systems in terms of whether the actual services rendered to the customers commensurate with their delay probabilities. To this end we employ the Resource Allocation Queue Fairness RAQF metrics to study and appraise “fairness” perceptions as indicator for service delivery quality and hence customers’ satisfaction in financial institution in Kaduna metropolis. Tentatively, we evaluate both the performance measures and the unfairness characteristics of the system, with respect to Poisson arrival process and exponentially distributed service time in M M 4 single queue system. Summarily, the results of the analyses show that, both the unfairness and the discrimination coefficients of the system in general increases with the probability of having more k m customers in the system and vice versa. The high negative discrimination index as well as its corresponding high unfairness coefficient, shows that though the ATM system may be necessary to improving service delivery quality, but queuing all customers with varying job size service requirement differences in a single queue structure served under FCFS job seniority service policy is unfair and hence may provoke customers’ dissatisfaction. To reverse the situation, the study recommended that i since all customers arriving at the system at any epoch have varying job size, then scheduling of services based on service requirement differences, and dedicating separate ATM machines to customers’ with similar job size will reduce the unfairness index and hence enhance customers’ satisfaction. ii To reduce the ATM overutilization factor during peak periods, as well as the high delay probability associated with the system, the deployment of few ATM machines with optimal processing speed would be more cost effective than more ATM machines with relatively slow processing speed and iii Since the efficient utilization of the ATM machine is a function of good internet connectivity, steady electricity supply and good computer literacy level of customers, therefore, the provision of good internet connectivity, alternative steady electricity supply solar , as well as provision of basic ATM operation guides to customers would reduce customer’s sojourn time at the service point, reduce the system unfairness index and hence guarantee customer’s satisfaction.
ATM queuing system, System discrimination and unfairness index, RAQF metrics, FCFS service policy, Poisson arrival processes, service requirement differences
Israel Udoh | Eno-bong Ukoh | Harmony Igoru | Linus Okafor