ORAHS2024
Abstract Submission

184. Comparison of two optimization models for a nurse scheduling and routing problem under uncertainty

Contributed abstract in session FA-4: Home Care /2, stream Regular talks.

Friday, 9:20-10:30
Room: Room S3

Authors (first author is the speaker)

1. Paul Fleurance
IMT Atantique
2. Olivier Péton
Campus de Nantes, IMT Atlantique Bretagne-Pays de la Loire
3. Maria I. Restrepo
Automation, Production and Computer Science, IMT Atlantique

Abstract

In this work we study the combination of two complex problems: scheduling and routing of nurses in the context of Home Health Care (HHC) activities. We seek to determine an optimal planning and routing of a team of nurses by answering the following questions: which nurse is working on which day and during which hours? For each working day, what is the sequence of visits to the patients?
We consider a time horizon of several days, continuity of care and legal constraints (min/max working hours per day/week, min/max number of days off per week, min resting time between two working shifts and max number of night shifts in a row).
In addition, we consider two sources of uncertainty that, to the best of our knowledge, have not been yet considered together in these types of problems: uncertainty on the service time and uncertainty on nurse availability (a nurse can get sick in the morning and not being able to visit their patients).
We model the problem as a two-stage stochastic programming model, and compare the two following approaches: assigning nurses to patients in the first or in the second stage. The first model is expected to be less flexible but easier to solve (the second stage problems are travelling salesman problems). The second model is more flexible but harder to solve (the subproblems are vehicle routing problems). We solve both models using the L-shaped method and present our numerical results on a set of randomly generated instances.

Keywords

Status: accepted

Back to the list of papers