Optimal management of a service facility having cross trained servers with front and back room operations

Abstract

newlineProviding excellent service quality and customer satisfaction are newlinethe important issues in a service industry Service quality is important and is newlineunavoidable for competitiveness of a service industry In this thesis a service newlinefacility with front and back room has limited number of servers are newlineconsidered Managers should balance a reduction in customer waiting time newlinewith high service costs and possibly reduced server efficiency due to cross newlinetraining Allowing servers to be crosstrained to perform both functions newlinecreates a potentially high productivity system with server utilization rates newlinenearly 100 percent newlineQueueing models are effectively used to obtain the performance newlinemeasures of systems This thesis is concerned with a problem faced by a newlineservice facility with front room and back room operations The front room newlineservers directly serve the customers from a queue whereas the back room newlineservers perform the job generated by the front room such as sorting of newlinematerial processing paperwork and restocking of shelves This is simplified newlinemodel for service facilities such as banks retail stores post offices etc newlineA problem faced by a service facility with cross trained server and newlineoptional feedback are considered in this thesis In the first model upon newlinecompletion of service in the front room the customer either leaves the system newlineor if he is not satisfied with the service the customer may enter the buffer as a newlinefeedback customer After completing the back room service the job may quit newlinethe system forever or may enter the buffer as a feedback customer in the back newlineroom is discussed in the second model The objective is to minimize the mean newlinenumber of customer waiting in the queue subject to the back room constraint newlinefor each state of the system The models are represented as a Markovian newlinedecision problem and both the models are solved using linear programming newlinetechniques A numerical study is conducted to analyze the impact of various newlineparameters on the performance newlineA deterministic policy is implemented to find the customer waiting newlinetime in the queu