Scheduling Drivers for a Delivery Company by a Two-Stage Stochastic Programming Approach

University essay from KTH/Optimeringslära och systemteori

Author: Petra Hesselmar; [2017]

Keywords: ;

Abstract: In this thesis a study of scheduling drivers for a food-delivery company has been done. Historic data has been used to generate scenarios. Those scenarios are used in a two-stage stochastic optimization model. The first stage costs consists of the cost for scheduling a driver to a shift. The second stage cost is the cost for corrective decisions. If there are too few drivers scheduled the corrective decision is to call the more expensive drivers from staffing agencies. If there are too many scheduled drivers there will be punishments costs for having drivers not assigned to a route, called slack drivers. The scheduling is done so that there is a requirement of having a few drivers from staffing agencies each day and so that there is a limited amount of slack drivers allowed. The scheduling also consists of a punishment cost for each scheduled driver to avoid unnecessary scheduled drivers. The result shows how the cost changes when the requirement of how many staffing agency drivers are desired varies. The requirement of allowed slack drivers are also varied and the cost is evaluated for these changes. Finally the punishment cost for slack drivers and punishment cost for each hired driver are evaluated in a similar manner. The cost of interest is the cost corresponding to salaries paid out. The result indicates that the difference in cost for salaries is small when the requirements and punishment costs are chosen so that the total number of scheduled drivers are close to the expected number of routes given from the historic data.

  AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)