Optimal allocation of CT scan capacity among stochastic demand of different customer classes

This paper studies the problem of appointment scheduling for two CT scanners one of which typically provides service to emergency patients, whose demands are generated during the day and must be served as soon as possible. The other provides service to scheduled outpatients who arrive based on an appointment schedule with some randomness due to no-shows and nonscheduled inpatients which arrive at random. The capacity management problem facing the managers of CT-scanners is to dynamically decide which appointment requests to accept and how to schedule the three kinds of patients on the two pointed facilities to maximize total revenue. So the Markov Decision Process model was built and an appointment rule was put forward in this paper.