Simulation Modelling and Analysis of the Time Planners' Problem - Scheduling Patients for MRI Examinations
MetadataShow full item record
The goal of this thesis is to identify possible areas of improvement in the scheduling process for MRI appointments at St. Olavs University Hospital. Our attention has been on the time planning, as this is considered to be a bottleneck in the system. In absence of intelligent software, time planners manually schedule and reschedule patients for appointments. More formally, the problem to be solved in this project, referred to as the Time Planners' Problem, allocates a set of requests to a time and date and an MRI machine. It is an offline operational problem. The main contribution of this thesis is a simulation model for generating and scheduling 13 000 incoming requests each year. Additionally, the Time Planners' Problem is concerned with macro waiting time in appointment scheduling, a topic which have received relatively little attention in academia. A simulation model is developed where different scheduling strategies can be tested and evaluated. We have identified three scheduling rules that schedule requests one at a time, and two optimisation strategies that can schedule multiple requests simultaneously. The optimisation strategies are solved with the use of a Binary Integer Programme. The rules and strategies are first and foremost evaluated on the number of requests they are able to successfully schedule when we simulate more than a year's worth of requests. We are able to obtain useful insights into the scheduling process, which we translate into managerial insights for hospital decision makers. In particular, we suggest that the time planners get access to more intelligent software to aid the scheduling process. Some of the scheduling rules presented in this thesis are good candidates for further development into such software. We also suggest that the current capacity seems low in comparison with the demand, which can be mitigated by for example having at least on machine be open on Saturdays. By improving the scheduling process, we propose that less rescheduling of requests will be necessary, which is both time consuming for the time planners, and often result in unutilised time between appointments. The hospital can also become less dependent on expensive disruption management tools such as extending the opening hours of the MRI machines.